polygraf-ai/arxiv-acl-pubmed-abstracts-10k-sample

null

Analysis of the ABCA4 c.[2588G>C;5603A>T] Allele in the Australian Population.

null

Inherited retinal diseases (IRDs) are genetically and phenotypically diverse, and they cause significant morbidity worldwide. Importantly, IRDs may be amenable to precision medicine strategies, and thus the molecular characterisation of causative variants is becoming increasingly important with the promise of personalised therapies on the horizon. ABCA4, involved in the translocation of visual cycle derivatives, is a well-established, frequent cause of IRDs worldwide, with pathogenic variants implicated in phenotypically diverse diseases. Identification of causative ABCA4 variants in some individuals, however, has been enigmatic, and resolution of this issue is currently a hotbed of research. Recent evidence has indicated that hypomorphic alleles, which cause disease under certain conditions, may account for some of the missing causal variants. It has been postulated that the ABCA4 c.5603A>T (p.Asn1868Ile) variant, previously considered benign, be reclassified as hypomorphic when in cis configuration with c.2588G>C (p.Gly863Ala/Gly863del), a variant previously considered to be pathogenic in its own right. We are exploring this relationship within an Australian cohort to test this theory.

null

pubmed

null

2,019

null

31,884,623

Advances in experimental medicine and biology

null

Spatial variation of heavy metal contamination in the riparian sediments after two-year flow regulation in the Three Gorges Reservoir, China.

null

Regular impoundment of the Three Gorges Reservoir (TGR) with intensified human activities in the watershed imparts a significant effect on the environmental changes in the riparian zone. In this study, six heavy metals (Cd, Cr, Cu, Ni, Pb and Zn) in the riparian sediments of the entire TGR mainstream were investigated in 2014 and 2016 to identify their contamination and risk characteristics and decipher the main factors for the variation of the metal contamination. The results showed that the concentrations of the heavy metals in the sediments did not vary significantly between 2014 and 2016, and their contamination degrees decreased in the order of Cd> > Cu ≈ Zn > Pb > Cr ≈ Ni in 2014 and Cd> > Zn > Cu ≈ Pb > Cr ≈ Ni in 2016. The potential eco-risk of Cd was extremely high in the two years, while the eco-risk of other metals was very low. The sediments showed a moderate to high contamination level, a high potential eco-risk but a low toxic risk to aquatic biota in the two years. Spatially, the contamination and risk levels of heavy metals were relatively higher in the downstream TGR region in 2014 except for the sites close to the urban areas but in the upper-middle TGR region in 2016. Increasing anthropogenic influence contributed to the high contamination and risk levels of Cd, Cu, Pb and Zn in the upper-middle region in 2016. The results indicated that the Cd contamination in the riparian sediments of the TGR was still a vital environmental issue, and the decreased sediment inputs from the upstream major tributaries, the periodic and anti-seasonal flow regulation, local geomorphological characteristics and anthropogenic activities determined the contamination distribution of heavy metals in the riparian sediments.

null

pubmed

null

2,019

null

30,308,875

The Science of the total environment

null

Vector of Locally Aggregated Embeddings for Text Representation

null

We present Vector of Locally Aggregated Embeddings (VLAE) for effective and, ultimately, lossless representation of textual content. Our model encodes each input text by effectively identifying and integrating the representations of its semantically-relevant parts. The proposed model generates high quality representation of textual content and improves the classification performance of current state-of-the-art deep averaging networks across several text classification tasks.

null

acl

inproceedings

amiri-mohtarami-2019-vector

Burstein, Jill and Doran, Christy and Solorio, Thamar

jun

2,019

Minneapolis, Minnesota

Association for Computational Linguistics

https://aclanthology.org/N19-1143/

Amiri, Hadi and Mohtarami, Mitra

Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)

1408--1414

null

hep-lat/0108020

Keh-Fei Liu

S.J. Dong, T. Draper, I. Horvath, F.X. Lee, K.F. Liu, and J.B. Zhang

Chiral Properties of Pseudoscalar Mesons on a Quenched $20^4$ Lattice with Overlap Fermions

Version accepted for publication by PRD. A few minor typographical errors have been corrected. 24 pages, 11 figures

Phys.Rev.D65:054507,2002

10.1103/PhysRevD.65.054507

UK/01-04

hep-lat

null

The chiral properties of the pseudoscalar mesons are studied numerically on a quenched $20^4$ lattice with the overlap fermion. We elucidate the role of the zero modes in the meson propagators, particularly that of the pseudoscalar meson. The non-perturbative renormalization constant $Z_A$ is determined from the axial Ward identity and is found to be almost independent of the quark mass for the range of quark masses we study; this implies that the $O(a^2)$ error is small. The pion decay constant, $f_{\pi}$, is calculated from which we determine the lattice spacing to be 0.148 fm. We look for quenched chiral log in the pseudoscalar decay constants and the pseudoscalar masses and we find clear evidence for its presence. The chiral log parameter $\delta$ is determined to be in the range 0.15 -- 0.4 which is consistent with that predicted from quenched chiral perturbation theory.

[ { "created": "Thu, 16 Aug 2001 03:27:14 GMT", "version": "v1" }, { "created": "Tue, 2 Oct 2001 19:36:05 GMT", "version": "v2" }, { "created": "Wed, 28 Nov 2001 19:32:01 GMT", "version": "v3" } ]

2008-11-26

[ [ "Dong", "S. J.", "" ], [ "Draper", "T.", "" ], [ "Horvath", "I.", "" ], [ "Lee", "F. X.", "" ], [ "Liu", "K. F.", "" ], [ "Zhang", "J. B.", "" ] ]

Thu, 16 Aug 2001 03:27:14 GMT

2001-08-16T03:27:14Z

[ "physics" ]

arxiv

null

2010.00178

William Clark

William H. Clark IV, Steven Hauser, William C. Headley, and Alan J. Michaels

Training Data Augmentation for Deep Learning Radio Frequency Systems

18 pages, 11 figures, journal, to be published in Journal of Defense Modeling and Simulation, revisions include typos and other revisions for clarity

null

10.1177/1548512921991245

null

cs.LG eess.SP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Applications of machine learning are subject to three major components that contribute to the final performance metrics. Within the category of neural networks, and deep learning specifically, the first two are the architecture for the model being trained and the training approach used. This work focuses on the third component, the data used during training. The primary questions that arise are ``what is in the data'' and ``what within the data matters?'' Looking into the Radio Frequency Machine Learning (RFML) field of Automatic Modulation Classification (AMC) as an example of a tool used for situational awareness, the use of synthetic, captured, and augmented data are examined and compared to provide insights about the quantity and quality of the available data necessary to achieve desired performance levels. There are three questions discussed within this work: (1) how useful a synthetically trained system is expected to be when deployed without considering the environment within the synthesis, (2) how can augmentation be leveraged within the RFML domain, and lastly, (3) what impact knowledge of degradations to the signal caused by the transmission channel contributes to the performance of a system. In general, the examined data types each have useful contributions to a final application, but captured data germane to the intended use case will always provide more significant information and enable the greatest performance. Despite the benefit of captured data, the difficulties and costs that arise from live collection often make the quantity of data needed to achieve peak performance impractical. This paper helps quantify the balance between real and synthetic data, offering concrete examples where training data is parametrically varied in size and source.

[ { "created": "Thu, 1 Oct 2020 02:26:16 GMT", "version": "v1" }, { "created": "Mon, 19 Oct 2020 16:34:08 GMT", "version": "v2" }, { "created": "Thu, 19 Nov 2020 20:20:59 GMT", "version": "v3" }, { "created": "Mon, 4 Jan 2021 15:50:24 GMT", "version": "v4" } ]

2021-02-23

[ [ "Clark", "William H.", "IV" ], [ "Hauser", "Steven", "" ], [ "Headley", "William C.", "" ], [ "Michaels", "Alan J.", "" ] ]

Thu, 1 Oct 2020 02:26:16 GMT

2020-10-01T02:26:16Z

[ "cs", "eess" ]

arxiv

null

cond-mat/0408409

Zolt\'an Eisler

Zoltan Eisler, Janos Kertesz, Soon-Hyung Yook, Albert-Laszlo Barabasi

Multiscaling and non-universality in fluctuations of driven complex systems

7 pages, 5 figures, submitted to Europhyisics Letters (extended text)

Europhys. Lett., 69 (4), pp. 664-670 (2005)

10.1209/epl/i2004-10384-1

null

cond-mat.stat-mech cond-mat.dis-nn q-fin.ST

null

For many externally driven complex systems neither the noisy driving force, nor the internal dynamics are a priori known. Here we focus on systems for which the time dependent activity of a large number of components can be monitored, allowing us to separate each signal into a component attributed to the external driving force and one to the internal dynamics. We propose a formalism to capture the potential multiscaling in the fluctuations and apply it to the high frequency trading records of the New York Stock Exchange. We find that on the time scale of minutes the dynamics is governed by internal processes, while on a daily or longer scale the external factors dominate. This transition from internal to external dynamics induces systematic changes in the scaling exponents, offering direct evidence of non-universality in the system.

[ { "created": "Wed, 18 Aug 2004 18:00:12 GMT", "version": "v1" }, { "created": "Mon, 8 Nov 2004 11:45:44 GMT", "version": "v2" } ]

2008-12-02

[ [ "Eisler", "Zoltan", "" ], [ "Kertesz", "Janos", "" ], [ "Yook", "Soon-Hyung", "" ], [ "Barabasi", "Albert-Laszlo", "" ] ]

Wed, 18 Aug 2004 18:00:12 GMT

2004-08-18T18:00:12Z

[ "physics", "q-fin" ]

arxiv

null

2204.01737

Eike Petersen

Eike Petersen and Aasa Feragen and Maria Luise da Costa Zemsch and Anders Henriksen and Oskar Eiler Wiese Christensen and Melanie Ganz

Feature robustness and sex differences in medical imaging: a case study in MRI-based Alzheimer's disease detection

Accepted for presentation at MICCAI 2022

null

eess.IV cs.CV cs.LG

http://creativecommons.org/licenses/by/4.0/

Convolutional neural networks have enabled significant improvements in medical image-based diagnosis. It is, however, increasingly clear that these models are susceptible to performance degradation when facing spurious correlations and dataset shift, leading, e.g., to underperformance on underrepresented patient groups. In this paper, we compare two classification schemes on the ADNI MRI dataset: a simple logistic regression model using manually selected volumetric features, and a convolutional neural network trained on 3D MRI data. We assess the robustness of the trained models in the face of varying dataset splits, training set sex composition, and stage of disease. In contrast to earlier work in other imaging modalities, we do not observe a clear pattern of improved model performance for the majority group in the training dataset. Instead, while logistic regression is fully robust to dataset composition, we find that CNN performance is generally improved for both male and female subjects when including more female subjects in the training dataset. We hypothesize that this might be due to inherent differences in the pathology of the two sexes. Moreover, in our analysis, the logistic regression model outperforms the 3D CNN, emphasizing the utility of manual feature specification based on prior knowledge, and the need for more robust automatic feature selection.

[ { "created": "Mon, 4 Apr 2022 17:37:54 GMT", "version": "v1" }, { "created": "Tue, 12 Apr 2022 11:30:53 GMT", "version": "v2" }, { "created": "Thu, 14 Jul 2022 16:35:45 GMT", "version": "v3" } ]

2022-07-15

[ [ "Petersen", "Eike", "" ], [ "Feragen", "Aasa", "" ], [ "Zemsch", "Maria Luise da Costa", "" ], [ "Henriksen", "Anders", "" ], [ "Christensen", "Oskar Eiler Wiese", "" ], [ "Ganz", "Melanie", "" ] ]

Mon, 4 Apr 2022 17:37:54 GMT

2022-04-04T17:37:54Z

[ "cs", "eess" ]

arxiv

null

Regulation of the DNA damage response on male meiotic sex chromosomes.

null

During meiotic prophase in males, the sex chromosomes partially synapse to form the XY body, a unique structure that recruits proteins involved in the DNA damage response, which is believed to be important for silencing of the sex chromosomes. It remains elusive how the DNA damage response in the XY body is regulated. Here we show that H2AX-MDC1-RNF8 signaling, which is well characterized in somatic cells, is dispensable for the recruitment of proteins to the unsynapsed axes in the XY body. On the other hand, the DNA damage response that spreads over the sex chromosomes is largely similar to that in somatic cells. This analysis shows that there are important differences between the regulation of the DNA damage response at the XY body and at DNA damage sites in somatic cells.

null

pubmed

null

2,013

null

23,812,044

Nature communications

null

Litigation Analytics: Extracting and querying motions and orders from {US} federal courts

null

Legal litigation planning can benefit from statistics collected from past decisions made by judges. Information on the typical duration for a submitted motion, for example, can give valuable clues for developing a successful strategy. Such information is encoded in semi-structured documents called dockets. In order to extract and aggregate this information, we deployed various information extraction and machine learning techniques. The aggregated data can be queried in real time within the Westlaw Edge search engine. In addition to a keyword search for judges, lawyers, law firms, parties and courts, we also implemented a question answering interface that offers targeted questions in order to get to the respective answers quicker.

null

acl

inproceedings

vacek-etal-2019-litigation

Ammar, Waleed and Louis, Annie and Mostafazadeh, Nasrin

jun

2,019

Minneapolis, Minnesota

Association for Computational Linguistics

https://aclanthology.org/N19-4020/

Vacek, Thomas and Song, Dezhao and Molina-Salgado, Hugo and Teo, Ronald and Cowling, Conner and Schilder, Frank

Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics (Demonstrations)

116--121

null

Questioning the Validity of Summarization Datasets and Improving Their Factual Consistency

null

The topic of summarization evaluation has recently attracted a surge of attention due to the rapid development of abstractive summarization systems. However, the formulation of the task is rather ambiguous, neither the linguistic nor the natural language processing communities have succeeded in giving a mutually agreed-upon definition. Due to this lack of well-defined formulation, a large number of popular abstractive summarization datasets are constructed in a manner that neither guarantees validity nor meets one of the most essential criteria of summarization: factual consistency. In this paper, we address this issue by combining state-of-the-art factual consistency models to identify the problematic instances present in popular summarization datasets. We release SummFC, a filtered summarization dataset with improved factual consistency, and demonstrate that models trained on this dataset achieve improved performance in nearly all quality aspects. We argue that our dataset should become a valid benchmark for developing and evaluating summarization systems.

null

acl

inproceedings

guo-etal-2022-questioning

Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue

dec

2,022

Abu Dhabi, United Arab Emirates

Association for Computational Linguistics

https://aclanthology.org/2022.emnlp-main.386/

Guo, Yanzhu and Clavel, Chlo{\'e} and Kamal Eddine, Moussa and Vazirgiannis, Michalis

Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing

5716--5727

null

2004.08730

Jian Ma

Predicting MMSE Score from Finger-Tapping Measurement

11 pages, 4 figures, 2 tables

Proceedings of 2021 Chinese Intelligent Automation Conference. Lecture Notes in Electrical Engineering, vol 801

10.1007/978-981-16-6372-7_34

null

q-bio.QM cs.LG stat.AP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Dementia is a leading cause of diseases for the elderly. Early diagnosis is very important for the elderly living with dementias. In this paper, we propose a method for dementia diagnosis by predicting MMSE score from finger-tapping measurement with machine learning pipeline. Based on measurement of finger tapping movement, the pipeline is first to select finger-tapping attributes with copula entropy and then to predict MMSE score from the selected attributes with predictive models. Experiments on real world data show that the predictive models such developed present good prediction performance. As a byproduct, the associations between certain finger-tapping attributes ('Number of taps', 'Average of intervals', and 'Frequency of taps' of both hands of bimanual in-phase task) and MMSE score are discovered with copula entropy, which may be interpreted as the biological relationship between cognitive ability and motor ability and therefore makes the predictive models explainable. The selected finger-tapping attributes can be considered as dementia biomarkers.

[ { "created": "Sat, 18 Apr 2020 23:30:57 GMT", "version": "v1" }, { "created": "Tue, 16 Nov 2021 02:52:54 GMT", "version": "v2" } ]

2021-11-17

[ [ "Ma", "Jian", "" ] ]

Sat, 18 Apr 2020 23:30:57 GMT

2020-04-18T23:30:57Z

[ "q-bio", "cs", "stat" ]

arxiv

null

The {RWTH} {A}achen speech recognition and machine translation system for {IWSLT} 2012

null

In this paper, the automatic speech recognition (ASR) and statistical machine translation (SMT) systems of RWTH Aachen University developed for the evaluation campaign of the International Workshop on Spoken Language Translation (IWSLT) 2012 are presented. We participated in the ASR (English), MT (English-French, Arabic-English, Chinese-English, German-English) and SLT (English-French) tracks. For the MT track both hierarchical and phrase-based SMT decoders are applied. A number of different techniques are evaluated in the MT and SLT tracks, including domain adaptation via data selection, translation model interpolation, phrase training for hierarchical and phrase-based systems, additional reordering model, word class language model, various Arabic and Chinese segmentation methods, postprocessing of speech recognition output with an SMT system, and system combination. By application of these methods we can show considerable improvements over the respective baseline systems.

null

acl

inproceedings

peitz-etal-2012-rwth

null

dec # " 6-7"

2,012

Hong Kong, Table of contents

null

https://aclanthology.org/2012.iwslt-evaluation.7/

Peitz, Stephan and Mansour, Saab and Freitag, Markus and Feng, Minwei and Huck, Matthias and Wuebker, Joern and Nuhn, Malte and Nu{\ss}baum-Thom, Markus and Ney, Hermann

Proceedings of the 9th International Workshop on Spoken Language Translation: Evaluation Campaign

69--76

null

1304.2942

Damiano Brigo

Damiano Brigo, Giuseppe Di Graziano

Optimal execution comparison across risks and dynamics, with solutions for displaced diffusions

null

q-fin.TR q-fin.RM q-fin.ST

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We solve a version of the optimal trade execution problem when the mid asset price follows a displaced diffusion. Optimal strategies in the adapted class under various risk criteria, namely value-at-risk, expected shortfall and a new criterion called "squared asset expectation" (SAE), related to a version of the cost variance measure, are derived and compared. It is well known that displaced diffusions (DD) exhibit dynamics which are in-between arithmetic Brownian motions (ABM) and geometric Brownian motions (GBM) depending of the choice of the shift parameter. Furthermore, DD allows for changes in the support of the mid asset price distribution, allowing one to include a minimum permitted value for the mid price, either positive or negative. We study the dependence of the optimal solution on the choice of the risk aversion criterion. Optimal solutions across criteria and asset dynamics are comparable although differences are not negligible for high levels of risk aversion and low market impact assets. This is illustrated with numerical examples.

[ { "created": "Wed, 10 Apr 2013 12:56:50 GMT", "version": "v1" }, { "created": "Thu, 8 Aug 2013 16:46:00 GMT", "version": "v2" }, { "created": "Fri, 9 May 2014 14:13:15 GMT", "version": "v3" } ]

2014-05-12

[ [ "Brigo", "Damiano", "" ], [ "Di Graziano", "Giuseppe", "" ] ]

Wed, 10 Apr 2013 12:56:50 GMT

2013-04-10T12:56:50Z

[ "q-fin" ]

arxiv

null

cond-mat/0512538

Nobuki Maeda

T. Aoyama and N. Maeda

Induced Chern-Simons term of a paired electron state in the quantum Hall system

26 pages, 6 figures

Eur. Phys. J. B 54, 157 (2006)

10.1140/epjb/e2006-00436-x

EPHOU-05-008

cond-mat.mes-hall cond-mat.supr-con hep-th

null

The induced Chern-Simons term for a paired electron state is calculated in the quantum Hall system by using a field theory on the von Neumann lattice. The coefficient of the Chern-Simons term, which is the Hall conductance, has not only the usual term proportional to a filling factor due to P (parity) & T (time reversal) symmetry breaking but also correction terms due to P & T & U(1) symmetry breaking. The correction term essentially comes from the Nambu-Goldstone mode and depends on an infrared limit. It is shown that the correction term is related to a topological number of a gap function in the momentum space.

[ { "created": "Wed, 21 Dec 2005 13:52:11 GMT", "version": "v1" }, { "created": "Thu, 22 Dec 2005 11:35:25 GMT", "version": "v2" }, { "created": "Mon, 15 May 2006 13:46:19 GMT", "version": "v3" }, { "created": "Mon, 25 Dec 2006 06:50:52 GMT", "version": "v4" } ]

2009-11-11

[ [ "Aoyama", "T.", "" ], [ "Maeda", "N.", "" ] ]

Wed, 21 Dec 2005 13:52:11 GMT

2005-12-21T13:52:11Z

[ "physics" ]

arxiv

null

Chromium (VI) induced phytotoxicity and oxidative stress in pea (Pisum sativum L.): biochemical changes and translocation of essential nutrients.

null

Due to widespread industrial use, chromium (Cr) is considered a hazardous environmental pollutant. It is known to inhibit plant growth and development. The present study provides the evidence of the phytotoxicity of this metal on the pea (Pisum sativum L. cv Azad) plants. The plants of pea (Pisum sativum L.) were grown in refined sand under different concentrations i.e. 0.05, 0.1, 0.2, 0.3 and 0.4 mM of Cr (VI) in order to study the effect on growth and yield, photosynthetic pigments, relative water content, non-reducing sugar and protein with activity of certain enzymes like catalase, peroxidase, starch phosphorylase and ribonuclease. The analysis of the results showed that photosynthetic pigments (68.68%), relative water contents (62.77%), non-reducing sugar (66.66%) and protein (81.57%) were decrease along with reduction in plant height (52.69%) and leaf area (50.81%) of the pea plants. However, in response to various concentration of Cr exposed plants showed significant induction of reducing and total sugars with enzymes like catalase, starch phosphorylase and ribonuclease. The translocation of Cr in various part of pea plant have been found in order of root> stem> leaves>seeds which ranged between 34.8 to 217.3 mg g(-1) d.wt. (dry weight) in roots, 6.5 to 173.13 mg g(-1) d.wt. in shoot, 4.2 to 74.43 mg g(-1) d.wt. in leaves and 0.94 to 8.64 mg g(-1) d.wt. in seeds, that is also reflected by the transfer factor of Cr from refined sand to tested species.

null

pubmed

null

2,009

null

20,120,464

Journal of environmental biology

1906.04150

Huazhen Fang

Ning Tian and Huazhen Fang and Jian Chen and Yebin Wang

Nonlinear Double-Capacitor Model for Rechargeable Batteries: Modeling, Identification and Validation

null

IEEE Transactions on Control Systems Technology, 2020

10.1109/TCST.2020.2976036

null

eess.SY cs.SY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper proposes a new equivalent circuit model for rechargeable batteries by modifying a double-capacitor model proposed in [1]. It is known that the original model can address the rate capacity effect and energy recovery effect inherent to batteries better than other models. However, it is a purely linear model and includes no representation of a battery's nonlinear phenomena. Hence, this work transforms the original model by introducing a nonlinear-mapping-based voltage source and a serial RC circuit. The modification is justified by an analogy with the single-particle model. Two parameter estimation approaches, termed 1.0 and 2.0, are designed for the new model to deal with the scenarios of constant-current and variable-current charging/discharging, respectively. In particular, the 2.0 approach proposes the notion of Wiener system identification based on maximum a posteriori estimation, which allows all the parameters to be estimated in one shot while overcoming the nonconvexity or local minima issue to obtain physically reasonable estimates. An extensive experimental evaluation shows that the proposed model offers excellent accuracy and predictive capability. A comparison against the Rint and Thevenin models further points to its superiority. With high fidelity and low mathematical complexity, this model is beneficial for various real-time battery management applications.

[ { "created": "Mon, 10 Jun 2019 17:43:26 GMT", "version": "v1" }, { "created": "Mon, 9 Mar 2020 04:33:30 GMT", "version": "v2" }, { "created": "Thu, 12 Mar 2020 22:58:20 GMT", "version": "v3" } ]

2024-09-23

[ [ "Tian", "Ning", "" ], [ "Fang", "Huazhen", "" ], [ "Chen", "Jian", "" ], [ "Wang", "Yebin", "" ] ]

Mon, 10 Jun 2019 17:43:26 GMT

2019-06-10T17:43:26Z

[ "cs", "eess" ]

arxiv

null

2005.04544

Baihan Lin

Baihan Lin, Guillermo Cecchi, Djallel Bouneffouf, Jenna Reinen, Irina Rish

Unified Models of Human Behavioral Agents in Bandits, Contextual Bandits and RL

Proceeding of HBAI 2020. This article supersedes and extends our work arXiv:1706.02897 (MAB) and arXiv:1906.11286 (RL) into the Contextual Bandit (CB) framework. It generalized extensively into multi-armed bandits, contextual bandits and RL settings to create a unified framework of human behavioral agents

In Human Brain and Artificial Intelligence (pp. 14-33). Springer 2021

null

cs.AI cs.LG q-bio.NC stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Artificial behavioral agents are often evaluated based on their consistent behaviors and performance to take sequential actions in an environment to maximize some notion of cumulative reward. However, human decision making in real life usually involves different strategies and behavioral trajectories that lead to the same empirical outcome. Motivated by clinical literature of a wide range of neurological and psychiatric disorders, we propose here a more general and flexible parametric framework for sequential decision making that involves a two-stream reward processing mechanism. We demonstrated that this framework is flexible and unified enough to incorporate a family of problems spanning multi-armed bandits (MAB), contextual bandits (CB) and reinforcement learning (RL), which decompose the sequential decision making process in different levels. Inspired by the known reward processing abnormalities of many mental disorders, our clinically-inspired agents demonstrated interesting behavioral trajectories and comparable performance on simulated tasks with particular reward distributions, a real-world dataset capturing human decision-making in gambling tasks, and the PacMan game across different reward stationarities in a lifelong learning setting.

[ { "created": "Sun, 10 May 2020 01:43:39 GMT", "version": "v1" }, { "created": "Tue, 12 May 2020 01:55:54 GMT", "version": "v2" }, { "created": "Thu, 10 Sep 2020 10:14:12 GMT", "version": "v3" }, { "created": "Mon, 14 Sep 2020 15:23:22 GMT", "version": "v4" }, { "created": "Mon, 27 Dec 2021 10:39:16 GMT", "version": "v5" } ]

2021-12-28

[ [ "Lin", "Baihan", "" ], [ "Cecchi", "Guillermo", "" ], [ "Bouneffouf", "Djallel", "" ], [ "Reinen", "Jenna", "" ], [ "Rish", "Irina", "" ] ]

Sun, 10 May 2020 01:43:39 GMT

2020-05-10T01:43:39Z

[ "q-bio", "cs", "stat" ]

arxiv

null

2210.13397

Christoph L\"uscher

Christoph L\"uscher and Mohammad Zeineldeen and Zijian Yang and Tina Raissi and Peter Vieting and Khai Le-Duc and Weiyue Wang and Ralf Schl\"uter and Hermann Ney

Development of Hybrid ASR Systems for Low Resource Medical Domain Conversational Telephone Speech

ASR System Paper for HYKIST project

null

cs.CL cs.SD eess.AS

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Language barriers present a great challenge in our increasingly connected and global world. Especially within the medical domain, e.g. hospital or emergency room, communication difficulties and delays may lead to malpractice and non-optimal patient care. In the HYKIST project, we consider patient-physician communication, more specifically between a German-speaking physician and an Arabic- or Vietnamese-speaking patient. Currently, a doctor can call the Triaphon service to get assistance from an interpreter in order to help facilitate communication. The HYKIST goal is to support the usually non-professional bilingual interpreter with an automatic speech translation system to improve patient care and help overcome language barriers. In this work, we present our ASR system development efforts for this conversational telephone speech translation task in the medical domain for two languages pairs, data collection, various acoustic model architectures and dialect-induced difficulties.

[ { "created": "Mon, 24 Oct 2022 16:49:19 GMT", "version": "v1" }, { "created": "Wed, 26 Oct 2022 18:21:38 GMT", "version": "v2" }, { "created": "Wed, 20 Sep 2023 08:08:47 GMT", "version": "v3" }, { "created": "Fri, 22 Sep 2023 10:15:15 GMT", "version": "v4" } ]

2023-09-25

[ [ "Lüscher", "Christoph", "" ], [ "Zeineldeen", "Mohammad", "" ], [ "Yang", "Zijian", "" ], [ "Raissi", "Tina", "" ], [ "Vieting", "Peter", "" ], [ "Le-Duc", "Khai", "" ], [ "Wang", "Weiyue", "" ], [ "Schlüter", "Ralf", "" ], [ "Ney", "Hermann", "" ] ]

Mon, 24 Oct 2022 16:49:19 GMT

2022-10-24T16:49:19Z

[ "cs", "eess" ]

arxiv

null

1002.0077

Alexander Verbovetsky

Joseph Krasil'shchik and Alexander Verbovetsky

Geometry of jet spaces and integrable systems

54 pages; v2-v6 : minor corrections

J.Geom.Phys.61:1633-1674,2011

10.1016/j.geomphys.2010.10.012

null

math.DG math-ph math.AP math.MP nlin.SI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

An overview of some recent results on the geometry of partial differential equations in application to integrable systems is given. Lagrangian and Hamiltonian formalism both in the free case (on the space of infinite jets) and with constraints (on a PDE) are discussed. Analogs of tangent and cotangent bundles to a differential equation are introduced and the variational Schouten bracket is defined. General theoretical constructions are illustrated by a series of examples.

[ { "created": "Sat, 30 Jan 2010 17:43:49 GMT", "version": "v1" }, { "created": "Thu, 11 Feb 2010 13:23:57 GMT", "version": "v2" }, { "created": "Fri, 9 Apr 2010 12:03:08 GMT", "version": "v3" }, { "created": "Tue, 14 Sep 2010 21:34:11 GMT", "version": "v4" }, { "created": "Mon, 15 Nov 2010 16:52:41 GMT", "version": "v5" }, { "created": "Tue, 18 Dec 2012 00:49:40 GMT", "version": "v6" } ]

2012-12-19

[ [ "Krasil'shchik", "Joseph", "" ], [ "Verbovetsky", "Alexander", "" ] ]

Sat, 30 Jan 2010 17:43:49 GMT

2010-01-30T17:43:49Z

[ "math", "nlin" ]

arxiv

null

Self-assembled tetra- and pentanuclear nickel(II) aggregates from phenoxido-based ligand -bound {Ni2} fragments: carboxylate bridge controlled structures.

null

Three different carboxylato bridges (R = C2H5, CF3, and PhCH2 in RCO2(-)) have been used to obtain the supramolecular aggregates [Ni5(μ-H2bpmp)2(μ3-OH)2(μ1,3-O2CC2H5)6]·2H2O·4DMF (1·2H2O·4DMF), [Ni4(μ3-H2bpmp)2(μ3-OH)2(μ1,3-O2CCF3)2](CF3CO2)2·H2O (2·H2O), and [Ni4(μ3-H2bpmp)2(μ3-OH)2(μ1,3-O2CCH2Ph)2](PhCH2CO2)2·4H2O (3·4H2O) (H3bpmp =2,6-bis-[(3-hydroxy-propylimino)-methyl]-4-methyl-phenol) from the hydroxido-bridged dinuclear motif [Ni2(μ-H2bpmp)(OH)](2+). These complexes have been characterized by X-ray crystallography and magnetic measurements. A change from propanoate group to trifluoroacetate and phenylaceate groups provided different course of cluster assembly based on Ni2(μ-H2bpmp)2 fragments. The {Ni5(μ3-OH)2(μ1,3-O2CC2H5)6}(2+) core in 1 contains five Ni(II) ions in an hourglass (pentanuclear vertex-shared double cubane) arrangement. These compounds are new examples of [Ni5] and [Ni4] complexes where aggregation of the building motifs are guided by the nature of the carboxylate anions, which allows an effective tuning of the self-aggregate process within same ligand environment. The study of the magnetic properties reveals that 1 exhibits an S = 3 ground state. Nevertheless, the magnetization increases above the expected saturation value of 6 μB at higher fields, because of the suppression of antiferromagnetic exchange between the central and peripheral Ni(II) ions. Complexes 2 and 3 exhibit ferromagnetic exchange interactions that result in the S = 4 ground state. Examination of AC magnetic susceptibility showed that complex 2 in finely ground form behaves as spin glass with the spin-freezing temperature of ∼5.5 K. This behavior was attributed to the collapse of the structure upon the loss of interstitial solvent. Such property was not observed for complex 3, in which the bulkier carboxylate ligands provide for a more robust crystal packing and larger separation between the [Ni4O4] clusters.

null

pubmed

null

2,013

null

24,295,223

Inorganic chemistry

nlin/0507042

Eleftherios Gkioulekas

Eleftherios Gkioulekas and Ka Kit Tung

Is the subdominant part of the energy spectrum due to downscale energy cascade hidden in quasi-geostrophic turbulence?

23 pages; accepted at Discrete and Continuous Dynamical Systems B; Major revision

Discrete Contin. Dyn. Syst. Ser. B 7, 293-314 , 2007

null

nlin.CD

null

In systems governing two-dimensional turbulence, surface quasi-geostrophic turbulence, (more generally $\alpha$-turbulence), two-layer quasi-geostrophic turbulence, etc., there often exist two conservative quadratic quantities, one ``energy''-like and one ``enstrophy''-like. In a finite inertial range there are in general two spectral fluxes, one associated with each conserved quantity. We derive here an inequality comparing the relative magnitudes of the ``energy'' and ``enstrophy'' fluxes for finite or infinitesimal dissipations, and for hyper or hypo viscosities. When this inequality is satisfied, as is the case of 2D turbulence,where the energy flux contribution to the energy spectrum is small, the subdominant part will be effectively hidden. In sQG turbulence, it is shown that the opposite is true: the downscale energy flux becomes the dominant contribution to the energy spectrum. A combination of these two behaviors appears to be the case in 2-layer QG turbulence, depending on the baroclinicity of the system.

[ { "created": "Thu, 21 Jul 2005 15:48:41 GMT", "version": "v1" }, { "created": "Wed, 8 Mar 2006 22:27:12 GMT", "version": "v2" }, { "created": "Sun, 24 Sep 2006 15:12:48 GMT", "version": "v3" } ]

2010-08-26

[ [ "Gkioulekas", "Eleftherios", "" ], [ "Tung", "Ka Kit", "" ] ]

Thu, 21 Jul 2005 15:48:41 GMT

2005-07-21T15:48:41Z

[ "nlin" ]

arxiv

null

2009.03884

Arthur Castello Branco de Oliveira Mr.

Arthur Castello B. de Oliveira, Milad Siami, Eduardo D. Sontag

Edge Selection in Bilinear Dynamical Networks

8 pages, 7 figures, technical note

null

eess.SY cs.SY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We develop some basic principles for the design and robustness analysis of a continuous-time bilinear dynamical network, where an attacker can manipulate the strength of the interconnections/edges between some of the agents/nodes. We formulate the edge protection optimization problem of picking a limited number of attack-free edges and minimizing the impact of the attack over the bilinear dynamical network.In particular, the $\mathcal{H}_2$-norm of bilinear systems is known to capture robustness and performance properties analogous to its linear counterpart and provides valuable insights for identifying which edges are most sensitive to attacks. The exact optimization problem is combinatorial in the number of edges, and brute-force approaches show poor scalability. However, we show that the $\mathcal{H}_2$-norm as a cost function is supermodular and, therefore, allows for efficient greedy approximations of the optimal solution. We illustrate and compare the effectiveness of our theoretical findings via numerical simulations.

[ { "created": "Tue, 8 Sep 2020 17:52:58 GMT", "version": "v1" }, { "created": "Mon, 19 Oct 2020 15:48:22 GMT", "version": "v2" }, { "created": "Mon, 20 Mar 2023 17:01:13 GMT", "version": "v3" } ]

2023-03-21

[ [ "de Oliveira", "Arthur Castello B.", "" ], [ "Siami", "Milad", "" ], [ "Sontag", "Eduardo D.", "" ] ]

Tue, 8 Sep 2020 17:52:58 GMT

2020-09-08T17:52:58Z

[ "cs", "eess" ]

arxiv

null

2009.07672

Ahmed Mohamed Hussain

Ahmed Mohamed Hussain, Gabriele Oligeri, and Thiemo Voigt

The Dark (and Bright) Side of IoT: Attacks and Countermeasures for Identifying Smart Home Devices and Services

15 pages, 7 figures. Accepted for the 9th International Symposium On Security And Privacy On Internet of Things (SPIoT), 2020

null

10.1007/978-3-030-68884-4_10

null

cs.CR cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present a new machine learning-based attack that exploits network patterns to detect the presence of smart IoT devices and running services in the WiFi radio spectrum. We perform an extensive measurement campaign of data collection, and we build up a model describing the traffic patterns characterizing three popular IoT smart home devices, i.e., Google Nest Mini, Amazon Echo, and Amazon Echo Dot. We prove that it is possible to detect and identify with overwhelming probability their presence and the services running by the aforementioned devices in a crowded WiFi scenario. This work proves that standard encryption techniques alone are not sufficient to protect the privacy of the end-user, since the network traffic itself exposes the presence of both the device and the associated service. While more work is required to prevent non-trusted third parties to detect and identify the user's devices, we introduce Eclipse, a technique to mitigate these types of attacks, which reshapes the traffic making the identification of the devices and the associated services similar to the random classification baseline.

[ { "created": "Wed, 16 Sep 2020 13:28:59 GMT", "version": "v1" }, { "created": "Sun, 20 Sep 2020 19:53:37 GMT", "version": "v2" }, { "created": "Wed, 4 Nov 2020 15:06:06 GMT", "version": "v3" }, { "created": "Sun, 25 Jul 2021 08:26:23 GMT", "version": "v4" } ]

2021-07-27

[ [ "Hussain", "Ahmed Mohamed", "" ], [ "Oligeri", "Gabriele", "" ], [ "Voigt", "Thiemo", "" ] ]

Wed, 16 Sep 2020 13:28:59 GMT

2020-09-16T13:28:59Z

[ "cs" ]

arxiv

null

2110.01219

Soojung Yang

Soojung Yang and Doyeong Hwang and Seul Lee and Seongok Ryu and Sung Ju Hwang

Hit and Lead Discovery with Explorative RL and Fragment-based Molecule Generation

To be published in NeurIPS 2021

null

q-bio.QM cs.AI cs.LG

http://creativecommons.org/licenses/by/4.0/

Recently, utilizing reinforcement learning (RL) to generate molecules with desired properties has been highlighted as a promising strategy for drug design. A molecular docking program - a physical simulation that estimates protein-small molecule binding affinity - can be an ideal reward scoring function for RL, as it is a straightforward proxy of the therapeutic potential. Still, two imminent challenges exist for this task. First, the models often fail to generate chemically realistic and pharmacochemically acceptable molecules. Second, the docking score optimization is a difficult exploration problem that involves many local optima and less smooth surfaces with respect to molecular structure. To tackle these challenges, we propose a novel RL framework that generates pharmacochemically acceptable molecules with large docking scores. Our method - Fragment-based generative RL with Explorative Experience replay for Drug design (FREED) - constrains the generated molecules to a realistic and qualified chemical space and effectively explores the space to find drugs by coupling our fragment-based generation method and a novel error-prioritized experience replay (PER). We also show that our model performs well on both de novo and scaffold-based schemes. Our model produces molecules of higher quality compared to existing methods while achieving state-of-the-art performance on two of three targets in terms of the docking scores of the generated molecules. We further show with ablation studies that our method, predictive error-PER (FREED(PE)), significantly improves the model performance.

[ { "created": "Mon, 4 Oct 2021 07:21:00 GMT", "version": "v1" }, { "created": "Tue, 5 Oct 2021 15:22:33 GMT", "version": "v2" }, { "created": "Wed, 27 Oct 2021 03:44:26 GMT", "version": "v3" } ]

2021-10-28

[ [ "Yang", "Soojung", "" ], [ "Hwang", "Doyeong", "" ], [ "Lee", "Seul", "" ], [ "Ryu", "Seongok", "" ], [ "Hwang", "Sung Ju", "" ] ]

Mon, 4 Oct 2021 07:21:00 GMT

2021-10-04T07:21:00Z

[ "q-bio", "cs" ]

arxiv

null

1803.00055

Ryan Marcus

Ryan Marcus, Olga Papaemmanouil

Deep Reinforcement Learning for Join Order Enumeration

null

aiDM@SIGMOD 2018 Proceedings of the First International Workshop on Exploiting Artificial Intelligence Techniques for Data Management

10.1145/3211954.3211957

null

cs.DB cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Join order selection plays a significant role in query performance. However, modern query optimizers typically employ static join enumeration algorithms that do not receive any feedback about the quality of the resulting plan. Hence, optimizers often repeatedly choose the same bad plan, as they do not have a mechanism for "learning from their mistakes". In this paper, we argue that existing deep reinforcement learning techniques can be applied to address this challenge. These techniques, powered by artificial neural networks, can automatically improve decision making by incorporating feedback from their successes and failures. Towards this goal, we present ReJOIN, a proof-of-concept join enumerator, and present preliminary results indicating that ReJOIN can match or outperform the PostgreSQL optimizer in terms of plan quality and join enumeration efficiency.

[ { "created": "Wed, 28 Feb 2018 20:00:33 GMT", "version": "v1" }, { "created": "Mon, 12 Mar 2018 21:09:45 GMT", "version": "v2" } ]

2018-09-28

[ [ "Marcus", "Ryan", "" ], [ "Papaemmanouil", "Olga", "" ] ]

Wed, 28 Feb 2018 20:00:33 GMT

2018-02-28T20:00:33Z

[ "cs" ]

arxiv

null

2103.03808

Yutaka Hori

Mei Minami, Yuka Masumoto, Yoshihiro Okawa, Tomotake Sasaki, Yutaka Hori

Two-step reinforcement learning for model-free redesign of nonlinear optimal regulator

null

SICE Journal of Control, Measurement, and System Integration, vol. 16, no. 1, pp. 349--362, 2023

10.1080/18824889.2023.2278753

null

eess.SY cs.LG cs.SY

http://creativecommons.org/licenses/by/4.0/

In many practical control applications, the performance level of a closed-loop system degrades over time due to the change of plant characteristics. Thus, there is a strong need for redesigning a controller without going through the system modeling process, which is often difficult for closed-loop systems. Reinforcement learning (RL) is one of the promising approaches that enable model-free redesign of optimal controllers for nonlinear dynamical systems based only on the measurement of the closed-loop system. However, the learning process of RL usually requires a considerable number of trial-and-error experiments using the poorly controlled system that may accumulate wear on the plant. To overcome this limitation, we propose a model-free two-step design approach that improves the transient learning performance of RL in an optimal regulator redesign problem for unknown nonlinear systems. Specifically, we first design a linear control law that attains some degree of control performance in a model-free manner, and then, train the nonlinear optimal control law with online RL by using the designed linear control law in parallel. We introduce an offline RL algorithm for the design of the linear control law and theoretically guarantee its convergence to the LQR controller under mild assumptions. Numerical simulations show that the proposed approach improves the transient learning performance and efficiency in hyperparameter tuning of RL.

[ { "created": "Fri, 5 Mar 2021 17:12:33 GMT", "version": "v1" }, { "created": "Sun, 11 Sep 2022 07:10:25 GMT", "version": "v2" }, { "created": "Mon, 24 Apr 2023 14:00:39 GMT", "version": "v3" }, { "created": "Thu, 30 Nov 2023 18:07:34 GMT", "version": "v4" } ]

2023-12-01

[ [ "Minami", "Mei", "" ], [ "Masumoto", "Yuka", "" ], [ "Okawa", "Yoshihiro", "" ], [ "Sasaki", "Tomotake", "" ], [ "Hori", "Yutaka", "" ] ]

Fri, 5 Mar 2021 17:12:33 GMT

2021-03-05T17:12:33Z

[ "cs", "eess" ]

arxiv

null

1910.13253

Jun Wang

Max W. Y. Lam, Jun Wang, Dan Su, Dong Yu

Mixup-breakdown: a consistency training method for improving generalization of speech separation models

Accepted in a Lesson session in ICASSP2020

null

eess.AS cs.LG cs.SD

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Deep-learning based speech separation models confront poor generalization problem that even the state-of-the-art models could abruptly fail when evaluating them in mismatch conditions. To address this problem, we propose an easy-to-implement yet effective consistency based semi-supervised learning (SSL) approach, namely Mixup-Breakdown training (MBT). It learns a teacher model to "breakdown" unlabeled inputs, and the estimated separations are interpolated to produce more useful pseudo "mixup" input-output pairs, on which the consistency regularization could apply for learning a student model. In our experiment, we evaluate MBT under various conditions with ascending degrees of mismatch, including unseen interfering speech, noise, and music, and compare MBT's generalization capability against state-of-the-art supervised learning and SSL approaches. The result indicates that MBT significantly outperforms several strong baselines with up to 13.77% relative SI-SNRi improvement. Moreover, MBT only adds negligible computational overhead to standard training schemes.

[ { "created": "Mon, 28 Oct 2019 03:34:42 GMT", "version": "v1" }, { "created": "Mon, 25 Nov 2019 11:34:34 GMT", "version": "v2" }, { "created": "Tue, 3 Mar 2020 10:00:06 GMT", "version": "v3" } ]

2020-03-04

[ [ "Lam", "Max W. Y.", "" ], [ "Wang", "Jun", "" ], [ "Su", "Dan", "" ], [ "Yu", "Dong", "" ] ]

Mon, 28 Oct 2019 03:34:42 GMT

2019-10-28T03:34:42Z

[ "cs", "eess" ]

arxiv

null

1411.0733

Silvia Grigolon

Silvia Grigolon, Peter Sollich, Olivier C. Martin

Modeling the emergence of polarity patterns for the intercellular transport of auxin in plants

17 pages and 9 figures (Main Text), 9 pages and 4 figures (Supplementary Material), revised version with some rearrangements

J. R. Soc. Interface, 2015, 12, 20141223

10.1098/rsif.2014.1223

null

q-bio.TO cond-mat.stat-mech

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The hormone auxin is actively transported throughout plants via protein machineries including the dedicated transporter known as PIN. The associated transport is ordered with nearby cells driving auxin flux in similar directions. Here we provide a model of both the auxin transport and of the dynamics of cellular polarisation based on flux sensing. Our main findings are: (i) spontaneous intracellular PIN polarisation arises if PIN recycling dynamics are sufficiently non-linear, (ii) there is no need for an auxin concentration gradient, and (iii) ordered multi-cellular patterns of PIN polarisation are favored by molecular noise.

[ { "created": "Mon, 3 Nov 2014 23:22:22 GMT", "version": "v1" }, { "created": "Fri, 13 Mar 2015 11:05:33 GMT", "version": "v2" } ]

2016-04-12

[ [ "Grigolon", "Silvia", "" ], [ "Sollich", "Peter", "" ], [ "Martin", "Olivier C.", "" ] ]

Mon, 3 Nov 2014 23:22:22 GMT

2014-11-03T23:22:22Z

[ "q-bio", "physics" ]

arxiv

null

1703.06594

Navid Constantinou

Navid C. Constantinou

A barotropic model of eddy saturation

14 pages, 8 figures, submitted to the Journal of Physical Oceanography

J. Phys. Oceanogr., vol. 48(2), pp. 397-411, 2018

10.1175/JPO-D-17-0182.1

null

physics.ao-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

"Eddy saturation" refers to a regime in which the total volume transport of an oceanic current is insensitive to the wind stress strength. Baroclinicity is currently believed to be key to the development of an eddy-saturated state. In this paper, it is shown that eddy saturation can also occur in a purely barotropic flow over topography, without baroclinicity. Thus, eddy saturation is a fundamental property of barotropic dynamics above topography. It is demonstrated that the main factor controlling the appearance or not of eddy-saturated states in the barotropic setting is the structure of geostrophic contours, that is the contours of $f/H$ of the ratio of the Coriolis parameter to the ocean's depth. Eddy-saturated states occur when the geostrophic contours are open, that is when the geostrophic contours span the whole zonal extent of the domain. This minimal requirement for eddy-saturated states is demonstrated using numerical integrations of a single-layer quasi-geostrophic flow over two different topographies characterized by either open or closed geostrophic contours with parameter values loosely inspired by the Southern Ocean. In this setting, transient eddies are produced through a barotropic-topographic instability that occurs due tot the interaction of the large-scale zonal flow with the topography. Through the study of this barotropic-topographic instability insight is gained on how eddy-saturated states are established.

[ { "created": "Mon, 20 Mar 2017 04:40:40 GMT", "version": "v1" }, { "created": "Mon, 27 Mar 2017 17:37:56 GMT", "version": "v2" }, { "created": "Wed, 13 Sep 2017 13:57:22 GMT", "version": "v3" }, { "created": "Wed, 13 Dec 2017 06:34:08 GMT", "version": "v4" } ]

2018-02-22

[ [ "Constantinou", "Navid C.", "" ] ]

Mon, 20 Mar 2017 04:40:40 GMT

2017-03-20T04:40:40Z

[ "physics" ]

arxiv

null

2106.15910

Yuichi Tanaka

Masatoshi Nagahama, Koki Yamada, Yuichi Tanaka, Stanley H. Chan, Yonina C. Eldar

Graph Signal Restoration Using Nested Deep Algorithm Unrolling

null

10.1109/TSP.2022.3180546

null

eess.SP cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Graph signal processing is a ubiquitous task in many applications such as sensor, social, transportation and brain networks, point cloud processing, and graph neural networks. Often, graph signals are corrupted in the sensing process, thus requiring restoration. In this paper, we propose two graph signal restoration methods based on deep algorithm unrolling (DAU). First, we present a graph signal denoiser by unrolling iterations of the alternating direction method of multiplier (ADMM). We then suggest a general restoration method for linear degradation by unrolling iterations of Plug-and-Play ADMM (PnP-ADMM). In the second approach, the unrolled ADMM-based denoiser is incorporated as a submodule, leading to a nested DAU structure. The parameters in the proposed denoising/restoration methods are trainable in an end-to-end manner. Our approach is interpretable and keeps the number of parameters small since we only tune graph-independent regularization parameters. We overcome two main challenges in existing graph signal restoration methods: 1) limited performance of convex optimization algorithms due to fixed parameters which are often determined manually. 2) large number of parameters of graph neural networks that result in difficulty of training. Several experiments for graph signal denoising and interpolation are performed on synthetic and real-world data. The proposed methods show performance improvements over several existing techniques in terms of root mean squared error in both tasks.

[ { "created": "Wed, 30 Jun 2021 08:57:01 GMT", "version": "v1" }, { "created": "Sun, 26 Dec 2021 01:41:25 GMT", "version": "v2" }, { "created": "Thu, 2 Jun 2022 01:30:41 GMT", "version": "v3" } ]

2022-07-27

[ [ "Nagahama", "Masatoshi", "" ], [ "Yamada", "Koki", "" ], [ "Tanaka", "Yuichi", "" ], [ "Chan", "Stanley H.", "" ], [ "Eldar", "Yonina C.", "" ] ]

Wed, 30 Jun 2021 08:57:01 GMT

2021-06-30T08:57:01Z

[ "cs", "eess" ]

arxiv

null

1611.05385

Kouichi Takemura

Hikaru Igarashi and Kouichi Takemura

On two-parameter solutions of simultaneous ultradiscrete Painleve II equation with parity variables

39 pages, title changed, figures added, typos corrected

Journal of Mathematical Physics 59, 103502 (2018)

10.1063/1.5020867

null

math.CA math-ph math.MP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We introduce a simultaneous ultradiscrete Painleve II equation with parity variables, which is shown to be more suitable for studying two-parameter solutions than the single second-order ultradiscrete Painleve II equation with parity variables. We investigate several types of two-parameter solutions and the solutions which are related with the ultradiscrete limit of determinant type solutions of q-Painleve II equation.

[ { "created": "Wed, 16 Nov 2016 17:51:36 GMT", "version": "v1" }, { "created": "Tue, 4 Sep 2018 12:50:34 GMT", "version": "v2" } ]

2020-05-28

[ [ "Igarashi", "Hikaru", "" ], [ "Takemura", "Kouichi", "" ] ]

Wed, 16 Nov 2016 17:51:36 GMT

2016-11-16T17:51:36Z

[ "math" ]

arxiv

null

1808.07018

Ivana Bala\v{z}evi\'c

Ivana Bala\v{z}evi\'c, Carl Allen and Timothy M. Hospedales

Hypernetwork Knowledge Graph Embeddings

null

10.1007/978-3-030-30493-5_52

null

cs.LG stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Knowledge graphs are graphical representations of large databases of facts, which typically suffer from incompleteness. Inferring missing relations (links) between entities (nodes) is the task of link prediction. A recent state-of-the-art approach to link prediction, ConvE, implements a convolutional neural network to extract features from concatenated subject and relation vectors. Whilst results are impressive, the method is unintuitive and poorly understood. We propose a hypernetwork architecture that generates simplified relation-specific convolutional filters that (i) outperforms ConvE and all previous approaches across standard datasets; and (ii) can be framed as tensor factorization and thus set within a well established family of factorization models for link prediction. We thus demonstrate that convolution simply offers a convenient computational means of introducing sparsity and parameter tying to find an effective trade-off between non-linear expressiveness and the number of parameters to learn.

[ { "created": "Tue, 21 Aug 2018 17:05:28 GMT", "version": "v1" }, { "created": "Tue, 28 Aug 2018 17:50:05 GMT", "version": "v2" }, { "created": "Thu, 18 Oct 2018 14:49:00 GMT", "version": "v3" }, { "created": "Sun, 30 Jun 2019 17:49:55 GMT", "version": "v4" }, { "created": "Mon, 15 Jul 2019 11:22:00 GMT", "version": "v5" } ]

2019-09-12

[ [ "Balažević", "Ivana", "" ], [ "Allen", "Carl", "" ], [ "Hospedales", "Timothy M.", "" ] ]

Tue, 21 Aug 2018 17:05:28 GMT

2018-08-21T17:05:28Z

[ "cs", "stat" ]

arxiv

null

1604.06203

Judith Lehnert

S.A. Plotnikov, J. Lehnert, A.L. Fradkov, E. Sch\"oll

Synchronization in heterogeneous FitzHugh-Nagumo networks with hierarchical architecture

null

Phys. Rev. E 94, 012203 (2016)

10.1103/PhysRevE.94.012203

null

nlin.AO

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We study synchronization in heterogeneous FitzHugh-Nagumo networks. It is well known that heterogeneities in the nodes hinder synchronization when becoming too large. Here, we develop a controller to counteract the impact of these heterogeneities. We first analyze the stability of the equilibrium point in a ring network of heterogeneous nodes. We then derive a sufficient condition for synchronization in the absence of control. Based on these results we derive the controller providing synchronization for parameter values where synchronization without control is absent. We demonstrate our results in networks with different topologies. Particular attention is given to hierarchical (fractal) topologies, which are relevant for the architecture of the brain.

[ { "created": "Thu, 21 Apr 2016 07:44:01 GMT", "version": "v1" }, { "created": "Thu, 21 Jul 2016 09:27:44 GMT", "version": "v2" } ]

2016-08-10

[ [ "Plotnikov", "S. A.", "" ], [ "Lehnert", "J.", "" ], [ "Fradkov", "A. L.", "" ], [ "Schöll", "E.", "" ] ]

Thu, 21 Apr 2016 07:44:01 GMT

2016-04-21T07:44:01Z

[ "nlin" ]

arxiv

null

2006.02854

Christian Anti\'c

Analogical proportions

null

cs.LO cs.AI cs.LG cs.SC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Analogy-making is at the core of human and artificial intelligence and creativity with applications to such diverse tasks as proving mathematical theorems and building mathematical theories, common sense reasoning, learning, language acquisition, and story telling. This paper introduces from first principles an abstract algebraic framework of analogical proportions of the form `$a$ is to $b$ what $c$ is to $d$' in the general setting of universal algebra. This enables us to compare mathematical objects possibly across different domains in a uniform way which is crucial for AI-systems. It turns out that our notion of analogical proportions has appealing mathematical properties. As we construct our model from first principles using only elementary concepts of universal algebra, and since our model questions some basic properties of analogical proportions presupposed in the literature, to convince the reader of the plausibility of our model we show that it can be naturally embedded into first-order logic via model-theoretic types and prove from that perspective that analogical proportions are compatible with structure-preserving mappings. This provides conceptual evidence for its applicability. In a broader sense, this paper is a first step towards a theory of analogical reasoning and learning systems with potential applications to fundamental AI-problems like common sense reasoning and computational learning and creativity.

[ { "created": "Thu, 4 Jun 2020 13:44:36 GMT", "version": "v1" }, { "created": "Sat, 4 Dec 2021 16:24:42 GMT", "version": "v10" }, { "created": "Fri, 18 Feb 2022 17:16:25 GMT", "version": "v11" }, { "created": "Mon, 14 Mar 2022 17:29:07 GMT", "version": "v12" }, { "created": "Sun, 8 May 2022 12:15:52 GMT", "version": "v13" }, { "created": "Tue, 28 Feb 2023 16:47:25 GMT", "version": "v14" }, { "created": "Fri, 29 Dec 2023 11:01:45 GMT", "version": "v15" }, { "created": "Sun, 7 Jun 2020 13:54:42 GMT", "version": "v2" }, { "created": "Tue, 25 Aug 2020 14:30:38 GMT", "version": "v3" }, { "created": "Thu, 10 Dec 2020 14:52:31 GMT", "version": "v4" }, { "created": "Sat, 17 Apr 2021 14:36:37 GMT", "version": "v5" }, { "created": "Tue, 25 May 2021 12:12:56 GMT", "version": "v6" }, { "created": "Sun, 15 Aug 2021 14:21:56 GMT", "version": "v7" }, { "created": "Mon, 22 Nov 2021 20:59:26 GMT", "version": "v8" }, { "created": "Wed, 24 Nov 2021 21:50:43 GMT", "version": "v9" } ]

2024-01-01

[ [ "Antić", "Christian", "" ] ]

Thu, 4 Jun 2020 13:44:36 GMT

2020-06-04T13:44:36Z

[ "cs" ]

arxiv

null

math/0505102

Friedrich Knop

Friedrich Knop, Bart Van Steirteghem

Classification of smooth affine spherical varieties

v1: 23 pages, uses texdraw; v2: 25 pages, introduction updated, Lemma 7.2 fixed, references added, typos corrected

Transform. Groups 11 (2006) 495-516

10.1007/s00031-005-1116-3

null

math.AG math.RT math.SG

null

Let G be a complex reductive group. A normal G-variety X is called spherical if a Borel subgroup of G has a dense orbit in X. Of particular interest are spherical varieties which are smooth and affine since they form local models for multiplicity free Hamiltonian K-manifolds, K a maximal compact subgroup of G. In this paper, we classify all smooth affine spherical varieties up to coverings, central tori, and C*-fibrations.

[ { "created": "Fri, 6 May 2005 01:13:29 GMT", "version": "v1" }, { "created": "Thu, 4 Aug 2005 08:49:38 GMT", "version": "v2" } ]

2007-05-23

[ [ "Knop", "Friedrich", "" ], [ "Van Steirteghem", "Bart", "" ] ]

Fri, 6 May 2005 01:13:29 GMT

2005-05-06T01:13:29Z

[ "math" ]

arxiv

null

1907.07548

Santosh Kumar

Ayana Sarkar, Manuja Kothiyal, Santosh Kumar

Distribution of the ratio of two consecutive level spacings in orthogonal to unitary crossover ensembles

Published version

Phys. Rev. E 101, 012216 (2020)

10.1103/PhysRevE.101.012216

null

math-ph math.MP physics.data-an stat.AP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The ratio of two consecutive level spacings has emerged as a very useful metric in investigating universal features exhibited by complex spectra. It does not require the knowledge of density of states and is therefore quite convenient to compute in analyzing the spectrum of a general system. The Wigner-surmise-like results for the ratio distribution are known for the invariant classes of Gaussian random matrices. However, for the crossover ensembles, which are useful in modeling systems with partially broken symmetries, corresponding results have remained unavailable so far. In this work, we derive exact results for the distribution and average of the ratio of two consecutive level spacings in the Gaussian orthogonal to unitary crossover ensemble using a $3\times 3$ random matrix model. This crossover is useful in modeling time-reversal symmetry breaking in quantum chaotic systems. Although based on a $3\times 3$ matrix model, our results can also be applied in the study of large spectra, provided the symmetry-breaking parameter facilitating the crossover is suitably scaled. We substantiate this claim by considering Gaussian and Laguerre crossover ensembles comprising large matrices. Moreover, we apply our result to investigate the violation of time-reversal invariance in the quantum kicked rotor system.

[ { "created": "Wed, 17 Jul 2019 14:38:26 GMT", "version": "v1" }, { "created": "Sun, 2 Feb 2020 16:21:36 GMT", "version": "v2" } ]

2020-02-04

[ [ "Sarkar", "Ayana", "" ], [ "Kothiyal", "Manuja", "" ], [ "Kumar", "Santosh", "" ] ]

Wed, 17 Jul 2019 14:38:26 GMT

2019-07-17T14:38:26Z

[ "physics", "math", "stat" ]

arxiv

null

0706.0482

Constantinos Kardaras

Constantinos Kardaras and Gordan Zitkovic

Stability of the utility maximization problem with random endowment in incomplete markets

21 pages, revised version. To appear in "Mathematical Finance".

null

q-fin.PM math.OC math.PR

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We perform a stability analysis for the utility maximization problem in a general semimartingale model where both liquid and illiquid assets (random endowments) are present. Small misspecifications of preferences (as modeled via expected utility), as well as views of the world or the market model (as modeled via subjective probabilities) are considered. Simple sufficient conditions are given for the problem to be well-posed, in the sense the optimal wealth and the marginal utility-based prices are continuous functionals of preferences and probabilistic views.

[ { "created": "Mon, 4 Jun 2007 16:54:26 GMT", "version": "v1" }, { "created": "Thu, 28 Jun 2007 16:42:38 GMT", "version": "v2" }, { "created": "Tue, 16 Mar 2010 00:26:28 GMT", "version": "v3" } ]

2010-03-17

[ [ "Kardaras", "Constantinos", "" ], [ "Zitkovic", "Gordan", "" ] ]

Mon, 4 Jun 2007 16:54:26 GMT

2007-06-04T16:54:26Z

[ "math", "q-fin" ]

arxiv

null

2011.13320

Amil Khanzada

Gunvant Chaudhari, Xinyi Jiang, Ahmed Fakhry, Asriel Han, Jaclyn Xiao, Sabrina Shen, Amil Khanzada

Virufy: Global Applicability of Crowdsourced and Clinical Datasets for AI Detection of COVID-19 from Cough

null

cs.SD cs.LG eess.AS eess.SP

http://creativecommons.org/licenses/by/4.0/

Rapid and affordable methods of testing for COVID-19 infections are essential to reduce infection rates and prevent medical facilities from becoming overwhelmed. Current approaches of detecting COVID-19 require in-person testing with expensive kits that are not always easily accessible. This study demonstrates that crowdsourced cough audio samples recorded and acquired on smartphones from around the world can be used to develop an AI-based method that accurately predicts COVID-19 infection with an ROC-AUC of 77.1% (75.2%-78.3%). Furthermore, we show that our method is able to generalize to crowdsourced audio samples from Latin America and clinical samples from South Asia, without further training using the specific samples from those regions. As more crowdsourced data is collected, further development can be implemented using various respiratory audio samples to create a cough analysis-based machine learning (ML) solution for COVID-19 detection that can likely generalize globally to all demographic groups in both clinical and non-clinical settings.

[ { "created": "Thu, 26 Nov 2020 14:38:19 GMT", "version": "v1" }, { "created": "Wed, 2 Dec 2020 17:00:46 GMT", "version": "v2" }, { "created": "Fri, 25 Dec 2020 15:08:57 GMT", "version": "v3" }, { "created": "Sat, 9 Jan 2021 05:23:19 GMT", "version": "v4" } ]

2021-01-12

[ [ "Chaudhari", "Gunvant", "" ], [ "Jiang", "Xinyi", "" ], [ "Fakhry", "Ahmed", "" ], [ "Han", "Asriel", "" ], [ "Xiao", "Jaclyn", "" ], [ "Shen", "Sabrina", "" ], [ "Khanzada", "Amil", "" ] ]

Thu, 26 Nov 2020 14:38:19 GMT

2020-11-26T14:38:19Z

[ "cs", "eess" ]

arxiv

null

2005.01578

Pedro Ricardo Ariel Salvador Bassi Electrical Engineer

Pedro R. A. S. Bassi, Romis Attux

A Deep Convolutional Neural Network for COVID-19 Detection Using Chest X-Rays

null

10.1007/s42600-021-00132-9

null

eess.IV cs.CV cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Purpose: We present image classifiers based on Dense Convolutional Networks and transfer learning to classify chest X-ray images according to three labels: COVID-19, pneumonia and normal. Methods: We fine-tuned neural networks pretrained on ImageNet and applied a twice transfer learning approach, using NIH ChestX-ray14 dataset as an intermediate step. We also suggested a novelty called output neuron keeping, which changes the twice transfer learning technique. In order to clarify the modus operandi of the models, we used Layer-wise Relevance Propagation (LRP) to generate heatmaps. Results: We were able to reach test accuracy of 100% on our test dataset. Twice transfer learning and output neuron keeping showed promising results improving performances, mainly in the beginning of the training process. Although LRP revealed that words on the X-rays can influence the networks' predictions, we discovered this had only a very small effect on accuracy. Conclusion: Although clinical studies and larger datasets are still needed to further ensure good generalization, the state-of-the-art performances we achieved show that, with the help of artificial intelligence, chest X-rays can become a cheap and accurate auxiliary method for COVID-19 diagnosis. Heatmaps generated by LRP improve the interpretability of the deep neural networks and indicate an analytical path for future research on diagnosis. Twice transfer learning with output neuron keeping improved performances.

[ { "created": "Thu, 30 Apr 2020 13:20:42 GMT", "version": "v1" }, { "created": "Thu, 18 Jun 2020 18:15:40 GMT", "version": "v2" }, { "created": "Fri, 11 Dec 2020 06:48:36 GMT", "version": "v3" }, { "created": "Wed, 13 Jan 2021 04:08:10 GMT", "version": "v4" } ]

2021-04-06

[ [ "Bassi", "Pedro R. A. S.", "" ], [ "Attux", "Romis", "" ] ]

Thu, 30 Apr 2020 13:20:42 GMT

2020-04-30T13:20:42Z

[ "cs", "eess" ]

arxiv

null

Unsupervised Text Summarization of Long Documents using Dependency-based Noun Phrases and Contextual Order Arrangement

null

Unsupervised extractive summarization has recently gained importance since it does not require labeled data. Among unsupervised methods, graph-based approaches have achieved outstanding results. These methods represent each document by a graph, with sentences as nodes and word-level similarity among sentences as edges. Common words can easily lead to a strong connection between sentence nodes. Thus, sentences with many common words can be misinterpreted as salient sentences for a summary. This work addresses the common word issue with a phrase-level graph that (1) focuses on the noun phrases of a document based on grammar dependencies and (2) initializes edge weights by term-frequency within the target document and inverse document frequency over the entire corpus. The importance scores of noun phrases extracted from the graph are then used to select the most salient sentences. To preserve summary coherence, the order of the selected sentences is re-arranged by a flow-aware orderBERT. The results reveal that our unsupervised framework outperformed other extractive methods on ROUGE as well as two human evaluations for semantic similarity and summary coherence.

null

acl

inproceedings

huang-etal-2022-unsupervised

Chang, Yung-Chun and Huang, Yi-Chin

nov

2,022

Taipei, Taiwan

The Association for Computational Linguistics and Chinese Language Processing (ACLCLP)

https://aclanthology.org/2022.rocling-1.3/

Huang, Yen-Hao and Lan, Hsiao-Yen and Chen, Yi-Shin

Proceedings of the 34th Conference on Computational Linguistics and Speech Processing (ROCLING 2022)

15--24

null

1101.1893

Carlos Gershenson

Rodrigo Poblanno-Balp and Carlos Gershenson

Modular Random Boolean Networks

33 pages, 14 figures, 11 tables. Corrected version, added experiments with large networks confirming results. Accepted in Artificial Life

Poblanno-Balp, R. & C. Gershenson (2011). Modular Random Boolean Networks. Artificial Life 17(4):331--351

10.1162/artl_a_00042

C3 Report 2011.01

nlin.CG nlin.AO q-bio.MN

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Random Boolean networks (RBNs) have been a popular model of genetic regulatory networks for more than four decades. However, most RBN studies have been made with random topologies, while real regulatory networks have been found to be modular. In this work, we extend classical RBNs to define modular RBNs. Statistical experiments and analytical results show that modularity has a strong effect on the properties of RBNs. In particular, modular RBNs have more attractors and are closer to criticality when chaotic dynamics would be expected, compared to classical RBNs.

[ { "created": "Mon, 10 Jan 2011 17:21:54 GMT", "version": "v1" }, { "created": "Fri, 18 Mar 2011 17:04:45 GMT", "version": "v2" } ]

2015-03-17

[ [ "Poblanno-Balp", "Rodrigo", "" ], [ "Gershenson", "Carlos", "" ] ]

Mon, 10 Jan 2011 17:21:54 GMT

2011-01-10T17:21:54Z

[ "q-bio", "nlin" ]

arxiv

null

2107.13405

Emanuele Lattanzi PhD.

Emanuele Lattanzi, Lorenzo Calisti, Valerio Freschi

Unstructured Handwashing Recognition using Smartwatch to Reduce Contact Transmission of Pathogens

null

IEEE Access, vol. 10, pp. 83111-83124, 2022

10.1109/ACCESS.2022.3197279

null

cs.LG cs.HC eess.SP

http://creativecommons.org/licenses/by-nc-nd/4.0/

Current guidelines from the World Health Organization indicate that the SARS-CoV-2 coronavirus, which results in the novel coronavirus disease (COVID-19), is transmitted through respiratory droplets or by contact. Contact transmission occurs when contaminated hands touch the mucous membrane of the mouth, nose, or eyes so hands hygiene is extremely important to prevent the spread of the SARSCoV-2 as well as of other pathogens. The vast proliferation of wearable devices, such as smartwatches, containing acceleration, rotation, magnetic field sensors, etc., together with the modern technologies of artificial intelligence, such as machine learning and more recently deep-learning, allow the development of accurate applications for recognition and classification of human activities such as: walking, climbing stairs, running, clapping, sitting, sleeping, etc. In this work, we evaluate the feasibility of a machine learning based system which, starting from inertial signals collected from wearable devices such as current smartwatches, recognizes when a subject is washing or rubbing its hands. Preliminary results, obtained over two different datasets, show a classification accuracy of about 95% and of about 94% for respectively deep and standard learning techniques.

[ { "created": "Wed, 28 Jul 2021 14:52:45 GMT", "version": "v1" }, { "created": "Mon, 15 Nov 2021 16:55:31 GMT", "version": "v2" }, { "created": "Wed, 9 Feb 2022 13:45:09 GMT", "version": "v3" }, { "created": "Mon, 6 Jun 2022 20:03:48 GMT", "version": "v4" } ]

2022-08-22

[ [ "Lattanzi", "Emanuele", "" ], [ "Calisti", "Lorenzo", "" ], [ "Freschi", "Valerio", "" ] ]

Wed, 28 Jul 2021 14:52:45 GMT

2021-07-28T14:52:45Z

[ "cs", "eess" ]

arxiv

null

1801.01584

Peter Pfaffelhuber

Peter Pfaffelhuber, Anton Wakolbinger

Fixation probabilities and hitting times under low levels of frequency-dependent selection

17 pages; This version includes a discussion section

null

math.PR q-bio.PE

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In population genetics, diffusions on the unit interval are often used to model the frequency path of an allele. In this setting we derive approximations for fixation probabilities, expected hitting times and the expected site-frequency-spectrum under small frequency-dependent selection. Specifically, we rederive and extend the one-third rule of evolutionary game theory (Nowak et al., 2004) and effects of stochastic slowdown (Altrock and Traulsen, 2009). Since similar effects are of interest in other application areas, we formulate our results for general one-dimensional diffusions.

[ { "created": "Thu, 4 Jan 2018 23:39:07 GMT", "version": "v1" }, { "created": "Wed, 18 Jul 2018 13:13:19 GMT", "version": "v2" }, { "created": "Thu, 6 Sep 2018 17:57:31 GMT", "version": "v3" }, { "created": "Mon, 17 Sep 2018 12:42:15 GMT", "version": "v4" } ]

2018-09-18

[ [ "Pfaffelhuber", "Peter", "" ], [ "Wakolbinger", "Anton", "" ] ]

Thu, 4 Jan 2018 23:39:07 GMT

2018-01-04T23:39:07Z

[ "q-bio", "math" ]

arxiv

null

2006.11942

Mehdi Abbana Bennani

Mehdi Abbana Bennani, Thang Doan, Masashi Sugiyama

Generalisation Guarantees for Continual Learning with Orthogonal Gradient Descent

null

stat.ML cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In Continual Learning settings, deep neural networks are prone to Catastrophic Forgetting. Orthogonal Gradient Descent was proposed to tackle the challenge. However, no theoretical guarantees have been proven yet. We present a theoretical framework to study Continual Learning algorithms in the Neural Tangent Kernel regime. This framework comprises closed form expression of the model through tasks and proxies for Transfer Learning, generalisation and tasks similarity. In this framework, we prove that OGD is robust to Catastrophic Forgetting then derive the first generalisation bound for SGD and OGD for Continual Learning. Finally, we study the limits of this framework in practice for OGD and highlight the importance of the Neural Tangent Kernel variation for Continual Learning with OGD.

[ { "created": "Sun, 21 Jun 2020 23:49:57 GMT", "version": "v1" }, { "created": "Mon, 6 Jul 2020 22:38:34 GMT", "version": "v2" }, { "created": "Mon, 16 Nov 2020 22:43:24 GMT", "version": "v3" }, { "created": "Fri, 4 Dec 2020 09:23:18 GMT", "version": "v4" } ]

2020-12-07

[ [ "Bennani", "Mehdi Abbana", "" ], [ "Doan", "Thang", "" ], [ "Sugiyama", "Masashi", "" ] ]

Sun, 21 Jun 2020 23:49:57 GMT

2020-06-21T23:49:57Z

[ "cs", "stat" ]

arxiv

null

2211.07767

Bo Dai

Hanjun Dai, Yuan Xue, Niao He, Bethany Wang, Na Li, Dale Schuurmans, Bo Dai

Learning to Optimize with Stochastic Dominance Constraints

Accepted to the 26th International Conference on Artificial Intelligence and Statistics (AISTATS 2023)

null

stat.ML cs.LG math.OC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In real-world decision-making, uncertainty is important yet difficult to handle. Stochastic dominance provides a theoretically sound approach for comparing uncertain quantities, but optimization with stochastic dominance constraints is often computationally expensive, which limits practical applicability. In this paper, we develop a simple yet efficient approach for the problem, the Light Stochastic Dominance Solver (light-SD), that leverages useful properties of the Lagrangian. We recast the inner optimization in the Lagrangian as a learning problem for surrogate approximation, which bypasses apparent intractability and leads to tractable updates or even closed-form solutions for gradient calculations. We prove convergence of the algorithm and test it empirically. The proposed light-SD demonstrates superior performance on several representative problems ranging from finance to supply chain management.

[ { "created": "Mon, 14 Nov 2022 21:54:31 GMT", "version": "v1" }, { "created": "Mon, 21 Nov 2022 18:15:24 GMT", "version": "v2" }, { "created": "Sat, 25 Feb 2023 00:04:59 GMT", "version": "v3" } ]

2023-02-28

[ [ "Dai", "Hanjun", "" ], [ "Xue", "Yuan", "" ], [ "He", "Niao", "" ], [ "Wang", "Bethany", "" ], [ "Li", "Na", "" ], [ "Schuurmans", "Dale", "" ], [ "Dai", "Bo", "" ] ]

Mon, 14 Nov 2022 21:54:31 GMT

2022-11-14T21:54:31Z

[ "cs", "math", "stat" ]

arxiv

null

1912.05189

Andr\'e Nortje

Andr\'e Nortje, Willie Brink, Herman A. Engelbrecht, Herman Kamper

BINet: a binary inpainting network for deep patch-based image compression

Signal Processing: Image Communication

Signal Processing: Image Communication 92C (2021) 116119

10.1016/j.image.2020.116119

null

eess.IV cs.CV

http://creativecommons.org/licenses/by-nc-nd/4.0/

Recent deep learning models outperform standard lossy image compression codecs. However, applying these models on a patch-by-patch basis requires that each image patch be encoded and decoded independently. The influence from adjacent patches is therefore lost, leading to block artefacts at low bitrates. We propose the Binary Inpainting Network (BINet), an autoencoder framework which incorporates binary inpainting to reinstate interdependencies between adjacent patches, for improved patch-based compression of still images. When decoding a patch, BINet additionally uses the binarised encodings from surrounding patches to guide its reconstruction. In contrast to sequential inpainting methods where patches are decoded based on previons reconstructions, BINet operates directly on the binary codes of surrounding patches without access to the original or reconstructed image data. Encoding and decoding can therefore be performed in parallel. We demonstrate that BINet improves the compression quality of a competitive deep image codec across a range of compression levels.

[ { "created": "Wed, 11 Dec 2019 09:11:01 GMT", "version": "v1" }, { "created": "Wed, 13 Jan 2021 16:33:53 GMT", "version": "v2" } ]

2021-01-14

[ [ "Nortje", "André", "" ], [ "Brink", "Willie", "" ], [ "Engelbrecht", "Herman A.", "" ], [ "Kamper", "Herman", "" ] ]

Wed, 11 Dec 2019 09:11:01 GMT

2019-12-11T09:11:01Z

[ "cs", "eess" ]

arxiv

null

Stem cells in dental pulp of deciduous teeth.

null

Dental pulp from deciduous (baby) teeth, which are discarded after exfoliation, represents an advantageous source of young stem cells. Herein, we discuss the methods of deciduous teeth stem cell (DTSC) isolation and cultivation. We show that based on these methods, at least three different stem cell populations can be identified: a population similar to bone marrow-derived mesenchymal stem cells, an epithelial stem-like cells, and/or a mixed population composed of both cell types. We analyzed the embryonic origin and stem cell niche of DTSCs with respect to the advantages they can provide for their future use in cell therapies and regenerative medicine. In vitro and in vivo differentiation of the DTSC populations, their developmental potential, immunological compatibility, tissue engineering, and transplantation use in studies in animal models are also the focus of the current report. We briefly describe the derivation of induced pluripotent stem (iPS) cells from DTSCs, which can be obtained more easily and efficiently in comparison with human fibroblasts. These iPS cells represent an interesting model for the investigation of pediatric diseases and disorders. The importance of DTSC banking is also discussed.

null

pubmed

null

2,012

null

22,032,258

Tissue engineering. Part B, Reviews

2211.12519

Lennart Bittel

Lennart Bittel, Sevag Gharibian, Martin Kliesch

Optimizing the depth of variational quantum algorithms is strongly QCMA-hard to approximate

31 pages, 2 figures

38th Computational Complexity Conference (CCC 2023), 34:1--34:24

10.4230/LIPIcs.CCC.2023.34

null

quant-ph cs.CC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Variational Quantum Algorithms (VQAs), such as the Quantum Approximate Optimization Algorithm (QAOA) of [Farhi, Goldstone, Gutmann, 2014], have seen intense study towards near-term applications on quantum hardware. A crucial parameter for VQAs is the \emph{depth} of the variational ``ansatz'' used -- the smaller the depth, the more amenable the ansatz is to near-term quantum hardware in that it gives the circuit a chance to be fully executed before the system decoheres. In this work, we show that approximating the optimal depth for a given VQA ansatz is intractable. Formally, we show that for any constant $\epsilon>0$, it is QCMA-hard to approximate the optimal depth of a VQA ansatz within multiplicative factor $N^{1-\epsilon}$, for $N$ denoting the encoding size of the VQA instance. (Here, Quantum Classical Merlin-Arthur (QCMA) is a quantum generalization of NP.) We then show that this hardness persists in the even ``simpler'' QAOA-type settings. To our knowledge, this yields the first natural QCMA-hard-to-approximate problems.

[ { "created": "Tue, 22 Nov 2022 19:00:01 GMT", "version": "v1" }, { "created": "Fri, 16 Jun 2023 07:47:31 GMT", "version": "v2" } ]

2023-07-12

[ [ "Bittel", "Lennart", "" ], [ "Gharibian", "Sevag", "" ], [ "Kliesch", "Martin", "" ] ]

Tue, 22 Nov 2022 19:00:01 GMT

2022-11-22T19:00:01Z

[ "physics", "cs" ]

arxiv

null

1911.12469

Kenji Shiohara

Koichi Miyamoto, Kenji Shiohara

Reduction of Qubits in Quantum Algorithm for Monte Carlo Simulation by Pseudo-random Number Generator

14 pages, 12 figures

Phys. Rev. A 102, 022424 (2020)

10.1103/PhysRevA.102.022424

EPHOU-19-0017

quant-ph q-fin.RM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

It is known that quantum computers can speed up Monte Carlo simulation compared to classical counterparts. There are already some proposals of application of the quantum algorithm to practical problems, including quantitative finance. In many problems in finance to which Monte Carlo simulation is applied, many random numbers are required to obtain one sample value of the integrand, since those problems are extremely high-dimensional integrations, for example, risk measurement of credit portfolio. This leads to the situation that the required qubit number is too large in the naive implementation where a quantum register is allocated per random number. In this paper, we point out that we can reduce qubits keeping quantum speed up if we perform calculation similar to classical one, that is, estimate the average of integrand values sampled by a pseudo-random number generator (PRNG) implemented on a quantum circuit. We present not only the overview of the idea but also concrete implementation of PRNG and application to credit risk measurement. Actually, reduction of qubits is a trade-off against increase of circuit depth. Therefore full reduction might be impractical, but such a trade-off between speed and memory space will be important in adjustment of calculation setting considering machine specs, if large-scale Monte Carlo simulation by quantum computer is in operation in the future.

[ { "created": "Thu, 28 Nov 2019 00:39:46 GMT", "version": "v1" }, { "created": "Fri, 28 Aug 2020 09:29:41 GMT", "version": "v2" } ]

2020-09-02

[ [ "Miyamoto", "Koichi", "" ], [ "Shiohara", "Kenji", "" ] ]

Thu, 28 Nov 2019 00:39:46 GMT

2019-11-28T00:39:46Z

[ "physics", "q-fin" ]

arxiv

null

1902.09807

Evgeny Tikhonov

S.U. Piatrusha, E.S. Tikhonov, Z.D. Kvon, N.N. Mikhailov, S.A. Dvoretsky, V.S. Khrapai

Topological protection brought to light by the time-reversal symmetry breaking

version accepted by Phys. Rev. Lett

Phys. Rev. Lett. 123, 056801 (2019)

10.1103/PhysRevLett.123.056801

null

cond-mat.mes-hall

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Recent topological band theory distinguishes electronic band insulators with respect to various symmetries and topological invariants, most commonly, the time reversal symmetry and the $\rm Z_2$ invariant. The interface of two topologically distinct insulators hosts a unique class of electronic states -- the helical states, which shortcut the gapped bulk and exhibit spin-momentum locking. The magic and so far elusive property of the helical electrons, known as topological protection, prevents them from coherent backscattering as long as the underlying symmetry is preserved. Here we present an experiment which brings to light the strength of topological protection in one-dimensional helical edge states of a $\rm Z_2$ quantum spin-Hall insulator in HgTe. At low temperatures, we observe the dramatic impact of a tiny magnetic field, which results in an exponential increase of the resistance accompanied by giant mesoscopic fluctuations and a gap opening. This textbook Anderson localization scenario emerges only upon the time-reversal symmetry breaking, bringing the first direct evidence of the topological protection strength in helical edge states.

[ { "created": "Tue, 26 Feb 2019 09:05:58 GMT", "version": "v1" }, { "created": "Fri, 12 Jul 2019 08:08:02 GMT", "version": "v2" } ]

2019-08-07

[ [ "Piatrusha", "S. U.", "" ], [ "Tikhonov", "E. S.", "" ], [ "Kvon", "Z. D.", "" ], [ "Mikhailov", "N. N.", "" ], [ "Dvoretsky", "S. A.", "" ], [ "Khrapai", "V. S.", "" ] ]

Tue, 26 Feb 2019 09:05:58 GMT

2019-02-26T09:05:58Z

[ "physics" ]

arxiv

null

Trace metals pollution in seawater and groundwater in the ship breaking area of Sitakund Upazilla, Chittagong, Bangladesh.

null

This study reveals potential accumulation of trace metals in the sea and groundwater due to ship breaking activities which take place along the Bay of Bengal in Sitakund Upazilla, Chittagong, Bangladesh. When compared with WHO and Bangladesh domestic standards for water quality, it is revealed that seawater was strongly polluted by Fe and Hg, moderately by Mn and Al, and slightly by Pb and Cd. Groundwater was strongly polluted by Fe, Pb and Hg, moderately by Mn and Al, and slightly by As. Trace element concentrations of all seawater samples exceeded the average concentration of elements in the Earth's seawater. The application of Principal Components Analysis identified two sources of pollution-marine and ship breaking. The mechanism of groundwater pollution inferred that if seawater is polluted, nearby groundwater is also polluted with trace metals due to the influence of seawater intrusion.

null

pubmed

null

2,013

null

23,465,568

Marine pollution bulletin

null

A dataset for identifying actionable feedback in collaborative software development

null

Software developers and testers have long struggled with how to elicit proactive responses from their coworkers when reviewing code for security vulnerabilities and errors. For a code review to be successful, it must not only identify potential problems but also elicit an active response from the colleague responsible for modifying the code. To understand the factors that contribute to this outcome, we analyze a novel dataset of more than one million code reviews for the Google Chromium project, from which we extract linguistic features of feedback that elicited responsive actions from coworkers. Using a manually-labeled subset of reviewer comments, we trained a highly accurate classifier to identify acted-upon comments (AUC = 0.85). Our results demonstrate the utility of our dataset, the feasibility of using NLP for this new task, and the potential of NLP to improve our understanding of how communications between colleagues can be authored to elicit positive, proactive responses.

null

acl

inproceedings

meyers-etal-2018-dataset

Gurevych, Iryna and Miyao, Yusuke

jul

2,018

Melbourne, Australia

Association for Computational Linguistics

https://aclanthology.org/P18-2021/

Meyers, Benjamin S. and Munaiah, Nuthan and Prud{'}hommeaux, Emily and Meneely, Andrew and Ovesdotter Alm, Cecilia and Wolff, Josephine and Murukannaiah, Pradeep K.

Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

126--131

null

Tregex and Tsurgeon: tools for querying and manipulating tree data structures

null

With syntactically annotated corpora becoming increasingly available for a variety of languages and grammatical frameworks, tree query tools have proven invaluable to linguists and computer scientists for both data exploration and corpus-based research. We provide a combined engine for tree query (Tregex) and manipulation (Tsurgeon) that can operate on arbitrary tree data structures with no need for preprocessing. Tregex remedies several expressive and implementational limitations of existing query tools, while Tsurgeon is to our knowledge the most expressive tree manipulation utility available.

null

acl

inproceedings

levy-andrew-2006-tregex

Calzolari, Nicoletta and Choukri, Khalid and Gangemi, Aldo and Maegaard, Bente and Mariani, Joseph and Odijk, Jan and Tapias, Daniel

may

2,006

Genoa, Italy

European Language Resources Association (ELRA)

https://aclanthology.org/L06-1311/

Levy, Roger and Andrew, Galen

Proceedings of the Fifth International Conference on Language Resources and Evaluation ({LREC}`06)

null

1509.00810

Gwynn Elfring

Charu Datt, Lailai Zhu, Gwynn J. Elfring, On Shun Pak

Squirming through shear-thinning fluids

null

JFM Rapids, 784 (2015) R1

10.1017/jfm.2015.600

null

physics.flu-dyn

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Many microorganisms find themselves immersed in fluids displaying non-Newtonian rheological properties such as viscoelasticity and shear-thinning viscosity. The effects of viscoelasticity on swimming at low Reynolds numbers have already received considerable attention, but much less is known about swimming in shear-thinning fluids. A general understanding of the fundamental question of how shear-thinning rheology influences swimming still remains elusive. To probe this question further, we study a spherical squirmer in a shear-thinning fluid using a combination of asymptotic analysis and numerical simulations. Shear-thinning rheology is found to affect a squirming swimmer in nontrivial and surprising ways; we predict and show instances of both faster and slower swimming depending on the surface actuation of the squirmer. We also illustrate that while a drag and thrust decomposition can provide insights into swimming in Newtonian fluids, extending this intuition to problems in complex media can prove problematic.

[ { "created": "Wed, 2 Sep 2015 18:19:21 GMT", "version": "v1" }, { "created": "Sat, 7 Nov 2015 21:08:21 GMT", "version": "v2" } ]

2015-11-10

[ [ "Datt", "Charu", "" ], [ "Zhu", "Lailai", "" ], [ "Elfring", "Gwynn J.", "" ], [ "Pak", "On Shun", "" ] ]

Wed, 2 Sep 2015 18:19:21 GMT

2015-09-02T18:19:21Z

[ "physics" ]

arxiv

null

2205.09054

Jo\~ao Morais

Jo\~ao Morais, Arash Behboodi, Hamed Pezeshki and Ahmed Alkhateeb

Position Aided Beam Prediction in the Real World: How Useful GPS Locations Actually Are?

Submitted to IEEE. Datasets and code files are available on the DeepSense website: https://deepsense6g.net/

null

eess.SP cs.IT cs.LG math.IT

http://creativecommons.org/licenses/by/4.0/

Millimeter-wave (mmWave) communication systems rely on narrow beams for achieving sufficient receive signal power. Adjusting these beams is typically associated with large training overhead, which becomes particularly critical for highly-mobile applications. Intuitively, since optimal beam selection can benefit from the knowledge of the positions of communication terminals, there has been increasing interest in leveraging position data to reduce the overhead in mmWave beam prediction. Prior work, however, studied this problem using only synthetic data that generally does not accurately represent real-world measurements. In this paper, we investigate position-aided beam prediction using a real-world large-scale dataset to derive insights into precisely how much overhead can be saved in practice. Furthermore, we analyze which machine learning algorithms perform best, what factors degrade inference performance in real data, and which machine learning metrics are more meaningful in capturing the actual communication system performance.

[ { "created": "Wed, 18 May 2022 16:39:17 GMT", "version": "v1" }, { "created": "Thu, 19 May 2022 22:25:27 GMT", "version": "v2" }, { "created": "Mon, 23 May 2022 06:04:01 GMT", "version": "v3" }, { "created": "Sun, 4 Sep 2022 23:35:47 GMT", "version": "v4" }, { "created": "Wed, 7 Sep 2022 19:13:43 GMT", "version": "v5" } ]

2022-09-09

[ [ "Morais", "João", "" ], [ "Behboodi", "Arash", "" ], [ "Pezeshki", "Hamed", "" ], [ "Alkhateeb", "Ahmed", "" ] ]

Wed, 18 May 2022 16:39:17 GMT

2022-05-18T16:39:17Z

[ "cs", "eess", "math" ]

arxiv

null

1911.10182

Jon Vadillo Jueguen

Jon Vadillo and Roberto Santana

Universal adversarial examples in speech command classification

14 pages, 2 figures, 4 tables; Revised external links

null

cs.LG eess.AS stat.ML

http://creativecommons.org/licenses/by-nc-sa/4.0/

Adversarial examples are inputs intentionally perturbed with the aim of forcing a machine learning model to produce a wrong prediction, while the changes are not easily detectable by a human. Although this topic has been intensively studied in the image domain, classification tasks in the audio domain have received less attention. In this paper we address the existence of universal perturbations for speech command classification. We provide evidence that universal attacks can be generated for speech command classification tasks, which are able to generalize across different models to a significant extent. Additionally, a novel analytical framework is proposed for the evaluation of universal perturbations under different levels of universality, demonstrating that the feasibility of generating effective perturbations decreases as the universality level increases. Finally, we propose a more detailed and rigorous framework to measure the amount of distortion introduced by the perturbations, demonstrating that the methods employed by convention are not realistic in audio-based problems.

[ { "created": "Fri, 22 Nov 2019 18:31:52 GMT", "version": "v1" }, { "created": "Tue, 26 Nov 2019 22:55:43 GMT", "version": "v2" }, { "created": "Sat, 21 Mar 2020 19:20:15 GMT", "version": "v3" }, { "created": "Sat, 13 Feb 2021 12:00:32 GMT", "version": "v4" } ]

2021-02-16

[ [ "Vadillo", "Jon", "" ], [ "Santana", "Roberto", "" ] ]

Fri, 22 Nov 2019 18:31:52 GMT

2019-11-22T18:31:52Z

[ "cs", "eess", "stat" ]

arxiv

null

2104.03139

Jun He Prof.

Han-Yu Jiang, Jun He

Functional annotation of creeping bentgrass protein sequences based on convolutional neural network

12 pages,3 figures. Accepted by BMC plant biology

BMC Plant Biol 22, 227 (2022)

10.1186/s12870-022-03607-8

null

q-bio.GN

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Background: Creeping bentgrass (Agrostis soionifera) is a perennial grass of Gramineae, belonging to cold season turfgrass, but has poor disease resistance. Up to now, little is known about the induced systemic resistance (ISR) mechanism, especially the relevant functional proteins, which is important to disease resistance of turfgrass. Achieving more information of proteins of infected creeping bentgrass is helpful to understand the ISR mechanism. Results: With BDO treatment, creeping bentgrass seedlings were grown, and the ISR response was induced by infecting Rhizoctonia solani. High-quality protein sequences of creeping bentgrass seedlings were obtained. Some of protein sequences were functionally annotated according to the database alignment while a large part of the obtained protein sequences was left non-annotated. To treat the non-annotated sequences, a prediction model based on convolutional neural network was established with the dataset from Uniport database in three domains to acquire good performance, especially the higher false positive control rate. With established model, the non-annotated protein sequences of creeping bentgrass were analyzed to annotate proteins relevant to disease-resistance response and signal transduction. Conclusions: The prediction model based on convolutional neural network was successfully applied to select good candidates of the proteins with functions relevant to the ISR mechanism from the protein sequences which cannot be annotated by database alignment. The waste of sequence data can be avoided, and research time and labor will be saved in further research of protein of creeping bentgrass by molecular biology technology. It also provides reference for other sequence analysis of turfgrass disease-resistance research.

[ { "created": "Wed, 7 Apr 2021 14:12:29 GMT", "version": "v1" }, { "created": "Tue, 24 May 2022 05:08:01 GMT", "version": "v2" } ]

2022-08-26

[ [ "Jiang", "Han-Yu", "" ], [ "He", "Jun", "" ] ]

Wed, 7 Apr 2021 14:12:29 GMT

2021-04-07T14:12:29Z

[ "q-bio" ]

arxiv

null

2108.04303

Jordan Awan

Jordan Awan and Salil Vadhan

Canonical Noise Distributions and Private Hypothesis Tests

23 pages + references and appendix. 4 figures

null

cs.CR math.ST stat.TH

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

$f$-DP has recently been proposed as a generalization of differential privacy allowing a lossless analysis of composition, post-processing, and privacy amplification via subsampling. In the setting of $f$-DP, we propose the concept of a canonical noise distribution (CND), the first mechanism designed for an arbitrary $f$-DP guarantee. The notion of CND captures whether an additive privacy mechanism perfectly matches the privacy guarantee of a given $f$. We prove that a CND always exists, and give a construction that produces a CND for any $f$. We show that private hypothesis tests are intimately related to CNDs, allowing for the release of private $p$-values at no additional privacy cost as well as the construction of uniformly most powerful (UMP) tests for binary data, within the general $f$-DP framework. We apply our techniques to the problem of difference of proportions testing, and construct a UMP unbiased (UMPU) "semi-private" test which upper bounds the performance of any $f$-DP test. Using this as a benchmark we propose a private test, based on the inversion of characteristic functions, which allows for optimal inference for the two population parameters and is nearly as powerful as the semi-private UMPU. When specialized to the case of $(\epsilon,0)$-DP, we show empirically that our proposed test is more powerful than any $(\epsilon/\sqrt 2)$-DP test and has more accurate type I errors than the classic normal approximation test.

[ { "created": "Mon, 9 Aug 2021 18:57:22 GMT", "version": "v1" }, { "created": "Wed, 8 Jun 2022 19:14:44 GMT", "version": "v2" }, { "created": "Mon, 13 Jun 2022 19:38:14 GMT", "version": "v3" }, { "created": "Fri, 13 Jan 2023 13:46:39 GMT", "version": "v4" } ]

2023-01-16

[ [ "Awan", "Jordan", "" ], [ "Vadhan", "Salil", "" ] ]

Mon, 9 Aug 2021 18:57:22 GMT

2021-08-09T18:57:22Z

[ "cs", "math", "stat" ]

arxiv

null

1106.2311

Jes\'us M. Mir\'o-Bueno

Jes\'us M. Mir\'o-Bueno and Alfonso Rodr\'iguez-Pat\'on

A simple negative interaction in the positive transcriptional feedback of a single gene is sufficient to produce reliable oscillations

25 pages, 12 figures, 3 tables

PLoS ONE 6(11): e27414. (2011)

10.1371/journal.pone.0027414

null

q-bio.MN q-bio.QM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Negative and positive transcriptional feedback loops are present in natural and synthetic genetic oscillators. A single gene with negative transcriptional feedback needs a time delay and sufficiently strong nonlinearity in the transmission of the feedback signal in order to produce biochemical rhythms. A single gene with only positive transcriptional feedback does not produce oscillations. Here, we demonstrate that this single-gene network in conjunction with a simple negative interaction can also easily produce rhythms. We examine a model comprised of two well-differentiated parts. The first is a positive feedback created by a protein that binds to the promoter of its own gene and activates the transcription. The second is a negative interaction in which a repressor molecule prevents this protein from binding to its promoter. A stochastic study shows that the system is robust to noise. A deterministic study identifies that the dynamics of the oscillator are mainly driven by two types of biomolecules: the protein, and the complex formed by the repressor and this protein. The main conclusion of this paper is that a simple and usual negative interaction, such as degradation, sequestration or inhibition, acting on the positive transcriptional feedback of a single gene is a sufficient condition to produce reliable oscillations. One gene is enough and the positive transcriptional feedback signal does not need to activate a second repressor gene. This means that at the genetic level an explicit negative feedback loop is not necessary. The model needs neither cooperative binding reactions nor the formation of protein multimers. Therefore, our findings could help to clarify the design principles of cellular clocks and constitute a new efficient tool for engineering synthetic genetic oscillators.

[ { "created": "Sun, 12 Jun 2011 14:10:56 GMT", "version": "v1" }, { "created": "Mon, 20 Jun 2011 18:48:18 GMT", "version": "v2" }, { "created": "Fri, 18 Nov 2011 19:49:50 GMT", "version": "v3" }, { "created": "Thu, 22 Dec 2011 15:51:43 GMT", "version": "v4" } ]

2011-12-23

[ [ "Miró-Bueno", "Jesús M.", "" ], [ "Rodríguez-Patón", "Alfonso", "" ] ]

Sun, 12 Jun 2011 14:10:56 GMT

2011-06-12T14:10:56Z

[ "q-bio" ]

arxiv

null

A Deeper Look into Sarcastic Tweets Using Deep Convolutional Neural Networks

null

Sarcasm detection is a key task for many natural language processing tasks. In sentiment analysis, for example, sarcasm can flip the polarity of an {\textquotedblleft}apparently positive{\textquotedblright} sentence and, hence, negatively affect polarity detection performance. To date, most approaches to sarcasm detection have treated the task primarily as a text categorization problem. Sarcasm, however, can be expressed in very subtle ways and requires a deeper understanding of natural language that standard text categorization techniques cannot grasp. In this work, we develop models based on a pre-trained convolutional neural network for extracting sentiment, emotion and personality features for sarcasm detection. Such features, along with the network`s baseline features, allow the proposed models to outperform the state of the art on benchmark datasets. We also address the often ignored generalizability issue of classifying data that have not been seen by the models at learning phase.

null

acl

inproceedings

poria-etal-2016-deeper

Matsumoto, Yuji and Prasad, Rashmi

dec

2,016

Osaka, Japan

The COLING 2016 Organizing Committee

https://aclanthology.org/C16-1151/

Poria, Soujanya and Cambria, Erik and Hazarika, Devamanyu and Vij, Prateek

Proceedings of {COLING} 2016, the 26th International Conference on Computational Linguistics: Technical Papers

1601--1612

null

2003.12150

Nuno Crokidakis

Modeling the early evolution of the COVID-19 in Brazil: results from a Susceptible-Infectious-Quarantined-Recovered (SIQR) model

8 pages, 2 figures, updated version (v2, with a different title) will appear in IJMPC

Int. J. Mod. Phys. C 31, 2050135 (2020)

10.1142/S0129183120501351

null

q-bio.PE physics.soc-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The world evolution of the Severe acute respiratory syndrome coronavirus 2 (SARS-Cov2 or simply COVID-19) led the World Health Organization to declare it a pandemic. The disease appeared in China in December 2019, and it has spread fast around the world, specially in european countries like Italy and Spain. The first reported case in Brazil was recorded in February 26, and after that the number of cases growed fast. In order to slow down the initial growth of the disease through the country, confirmed positive cases were isolated to not transmit the disease. To better understand the early evolution of COVID-19 in Brazil, we apply a Susceptible-Infectious-Quarantined-Recovered (SIQR) model to the analysis of data from the Brazilian Department of Health, obtained from February 26, 2020 through March 25, 2020. Based on analyical and numerical results, as well on the data, the basic reproduction number is estimated to $R_{0}=5.25$. In addition, we estimate that the ratio unidentified infectious individuals and confirmed cases at the beginning of the epidemic is about $10$, in agreement with previous studies. We also estimated the epidemic doubling time to be $2.72$ days.

[ { "created": "Thu, 26 Mar 2020 20:37:45 GMT", "version": "v1" }, { "created": "Fri, 22 May 2020 17:16:25 GMT", "version": "v2" } ]

2020-10-22

[ [ "Crokidakis", "Nuno", "" ] ]

Thu, 26 Mar 2020 20:37:45 GMT

2020-03-26T20:37:45Z

[ "q-bio", "physics" ]

arxiv

null

1811.12231

Robert Geirhos

Robert Geirhos, Patricia Rubisch, Claudio Michaelis, Matthias Bethge, Felix A. Wichmann, Wieland Brendel

ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness

Accepted at ICLR 2019 (oral)

null

cs.CV cs.AI cs.LG q-bio.NC stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Convolutional Neural Networks (CNNs) are commonly thought to recognise objects by learning increasingly complex representations of object shapes. Some recent studies suggest a more important role of image textures. We here put these conflicting hypotheses to a quantitative test by evaluating CNNs and human observers on images with a texture-shape cue conflict. We show that ImageNet-trained CNNs are strongly biased towards recognising textures rather than shapes, which is in stark contrast to human behavioural evidence and reveals fundamentally different classification strategies. We then demonstrate that the same standard architecture (ResNet-50) that learns a texture-based representation on ImageNet is able to learn a shape-based representation instead when trained on "Stylized-ImageNet", a stylized version of ImageNet. This provides a much better fit for human behavioural performance in our well-controlled psychophysical lab setting (nine experiments totalling 48,560 psychophysical trials across 97 observers) and comes with a number of unexpected emergent benefits such as improved object detection performance and previously unseen robustness towards a wide range of image distortions, highlighting advantages of a shape-based representation.

[ { "created": "Thu, 29 Nov 2018 15:04:05 GMT", "version": "v1" }, { "created": "Mon, 14 Jan 2019 13:59:09 GMT", "version": "v2" }, { "created": "Wed, 9 Nov 2022 23:15:15 GMT", "version": "v3" } ]

2022-11-11

[ [ "Geirhos", "Robert", "" ], [ "Rubisch", "Patricia", "" ], [ "Michaelis", "Claudio", "" ], [ "Bethge", "Matthias", "" ], [ "Wichmann", "Felix A.", "" ], [ "Brendel", "Wieland", "" ] ]

Thu, 29 Nov 2018 15:04:05 GMT

2018-11-29T15:04:05Z

[ "q-bio", "cs", "stat" ]

arxiv

null

2112.06180

Yueh-Cheng Liu

Bolivar Solarte, Yueh-Cheng Liu, Chin-Hsuan Wu, Yi-Hsuan Tsai, Min Sun

360-DFPE: Leveraging Monocular 360-Layouts for Direct Floor Plan Estimation

IEEE RA-L 2022

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present 360-DFPE, a sequential floor plan estimation method that directly takes 360-images as input without relying on active sensors or 3D information. Our approach leverages a loosely coupled integration between a monocular visual SLAM solution and a monocular 360-room layout approach, which estimate camera poses and layout geometries, respectively. Since our task is to sequentially capture the floor plan using monocular images, the entire scene structure, room instances, and room shapes are unknown. To tackle these challenges, we first handle the scale difference between visual odometry and layout geometry via formulating an entropy minimization process, which enables us to directly align 360-layouts without knowing the entire scene in advance. Second, to sequentially identify individual rooms, we propose a novel room identification algorithm that tracks every room along the camera exploration using geometry information. Lastly, to estimate the final shape of the room, we propose a shortest path algorithm with an iterative coarse-to-fine strategy, which improves prior formulations with higher accuracy and faster run-time. Moreover, we collect a new floor plan dataset with challenging large-scale scenes, providing both point clouds and sequential 360-image information. Experimental results show that our monocular solution achieves favorable performance against the current state-of-the-art algorithms that rely on active sensors and require the entire scene reconstruction data in advance.

[ { "created": "Sun, 12 Dec 2021 08:36:41 GMT", "version": "v1" }, { "created": "Tue, 21 Dec 2021 10:27:05 GMT", "version": "v2" }, { "created": "Sun, 8 May 2022 15:51:31 GMT", "version": "v3" } ]

2022-05-10

[ [ "Solarte", "Bolivar", "" ], [ "Liu", "Yueh-Cheng", "" ], [ "Wu", "Chin-Hsuan", "" ], [ "Tsai", "Yi-Hsuan", "" ], [ "Sun", "Min", "" ] ]

Sun, 12 Dec 2021 08:36:41 GMT

2021-12-12T08:36:41Z

[ "cs" ]

arxiv

null

2106.09367

Lukas Kades

Lukas Kades, Martin G\"arttner, Thomas Gasenzer, Jan M. Pawlowski

Towards sampling complex actions

29 pages, 9 figures

Phys. Rev. E 105, 045315 (2022)

10.1103/PhysRevE.105.045315

null

hep-lat hep-th math-ph math.MP stat.CO

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Path integrals with complex actions are encountered for many physical systems ranging from spin- or mass-imbalanced atomic gases and graphene to quantum chromo-dynamics at finite density to the non-equilibrium evolution of quantum systems. Many computational approaches have been developed for tackling the sign problem emerging for complex actions. Among these, complex Langevin dynamics has the appeal of general applicability. One of its key challenges is the potential convergence of the dynamics to unphysical fixed points. The statistical sampling process at such a fixed point is not based on the physical action and hence leads to wrong predictions. Moreover, its unphysical nature is hard to detect due to the implicit nature of the process. In the present work we set up a general approach based on a Markov chain Monte Carlo scheme in an extended state space. In this approach we derive an explicit real sampling process for generalized complex Langevin dynamics. Subject to a set of constraints, this sampling process is the physical one. These constraints originate from the detailed-balance equations satisfied by the Monte Carlo scheme. This allows us to re-derive complex Langevin dynamics from a new perspective and establishes a framework for the explicit construction of new sampling schemes for complex actions.

[ { "created": "Thu, 17 Jun 2021 10:37:43 GMT", "version": "v1" }, { "created": "Fri, 18 Jun 2021 10:52:43 GMT", "version": "v2" } ]

2022-09-01

[ [ "Kades", "Lukas", "" ], [ "Gärttner", "Martin", "" ], [ "Gasenzer", "Thomas", "" ], [ "Pawlowski", "Jan M.", "" ] ]

Thu, 17 Jun 2021 10:37:43 GMT

2021-06-17T10:37:43Z

[ "physics", "math", "stat" ]

arxiv

null

math/0108066

Wilberd van der Kallen

Cohomology with Grosshans graded coefficients

17 pages; correction of twelve subscripts

CRM Proceedings and Lecture Notes, Volume 35 (2004), 127-138

null

math.RT

null

Let the reductive group G act on the finitely generated commutative k-algebra A. We ask if the finite generation property of the ring of invariants A^G extends to the full rational cohomology ring H^*(G,A). We confirm this when G=SL_2 and also when the characteristic of k is large and the action on A is replaced by the `contracted' action on the Grosshans graded ring gr(A). In the printed version of this paper the first level of some spectral sequences was misidentified as the second. This affects twelve subscripts. We have corrected them (in 3.10, 4.11, 4.12).

[ { "created": "Thu, 9 Aug 2001 12:57:56 GMT", "version": "v1" }, { "created": "Mon, 17 Dec 2001 10:28:57 GMT", "version": "v2" }, { "created": "Fri, 3 Jan 2003 10:07:33 GMT", "version": "v3" }, { "created": "Thu, 24 Jun 2004 09:08:04 GMT", "version": "v4" } ]

2007-05-23

[ [ "van der Kallen", "Wilberd", "" ] ]

Thu, 9 Aug 2001 12:57:56 GMT

2001-08-09T12:57:56Z

[ "math" ]

arxiv

null

2106.15013

Dominik St\"oger

Dominik St\"oger and Mahdi Soltanolkotabi

Small random initialization is akin to spectral learning: Optimization and generalization guarantees for overparameterized low-rank matrix reconstruction

83 pages; published in NeurIPS 2021; corrected inaccuracies in the proofs of Lemma 9.5 and Theorem 9.6

null

cs.LG cs.IT math.IT math.OC math.ST stat.ML stat.TH

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Recently there has been significant theoretical progress on understanding the convergence and generalization of gradient-based methods on nonconvex losses with overparameterized models. Nevertheless, many aspects of optimization and generalization and in particular the critical role of small random initialization are not fully understood. In this paper, we take a step towards demystifying this role by proving that small random initialization followed by a few iterations of gradient descent behaves akin to popular spectral methods. We also show that this implicit spectral bias from small random initialization, which is provably more prominent for overparameterized models, also puts the gradient descent iterations on a particular trajectory towards solutions that are not only globally optimal but also generalize well. Concretely, we focus on the problem of reconstructing a low-rank matrix from a few measurements via a natural nonconvex formulation. In this setting, we show that the trajectory of the gradient descent iterations from small random initialization can be approximately decomposed into three phases: (I) a spectral or alignment phase where we show that that the iterates have an implicit spectral bias akin to spectral initialization allowing us to show that at the end of this phase the column space of the iterates and the underlying low-rank matrix are sufficiently aligned, (II) a saddle avoidance/refinement phase where we show that the trajectory of the gradient iterates moves away from certain degenerate saddle points, and (III) a local refinement phase where we show that after avoiding the saddles the iterates converge quickly to the underlying low-rank matrix. Underlying our analysis are insights for the analysis of overparameterized nonconvex optimization schemes that may have implications for computational problems beyond low-rank reconstruction.

[ { "created": "Mon, 28 Jun 2021 22:52:39 GMT", "version": "v1" }, { "created": "Fri, 5 Nov 2021 14:10:33 GMT", "version": "v2" }, { "created": "Thu, 18 Aug 2022 18:14:39 GMT", "version": "v3" }, { "created": "Mon, 26 Sep 2022 14:49:53 GMT", "version": "v4" } ]

2022-09-27

[ [ "Stöger", "Dominik", "" ], [ "Soltanolkotabi", "Mahdi", "" ] ]

Mon, 28 Jun 2021 22:52:39 GMT

2021-06-28T22:52:39Z

[ "cs", "math", "stat" ]

arxiv

null

A facile and efficient four-step enantioselective synthesis of (+)-vernolepin from (+)-minimolide, the major germacranolide of Mikania minima.

null

Enantiomerically pure (+)-vemolepin was semi-synthesized for the first time using the synthon (6S,7R,8S)-8,14-diacetoxy-15-hydroxygermacra-1(10),4,11(13)-trien-6, 12-olide [(+)-minimolide], the major sesquiterpene lactone of the Argentinean vine Mikania minima. After performing four consecutive reactions (Cope rearrangement, two oxidations, and selective hydrolysis of the acetate groups) on the synthon (+)-minimolide, a (+)-vernolepin yield of ca. 40% was achieved, proving to be a suitable semi-synthetic strategy for the production of quantities between 0.5-1.0 g of (+)-vernolepin. The transformations described here mimetize the biogenetic pathway for the production of (+)-vernolepin in the genus Vernonia. The synthesized (+)-vernolepin, but not its precursors, shows antifungal activity similar to amphotericin B. The semi-synthesis reported here combines affordable and easily available chemical reagents with classical organic methodologies.

null

pubmed

null

2,011

null

21,560,754

Natural product communications

1108.1607

Zhaoliang Wang

Zhaoliang Wang and Anmin Wang and Yang Yang and Xuechao Li

Exact Eigenfunctions of $N$-Body system with Quadratic Pair Potential

7 pages,1 figure

null

10.1088/0253-6102/58/5/04

null

math-ph cond-mat.quant-gas math.MP nucl-th quant-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We obtain all the exact eigenvalues and the corresponding eigenfunctions of $N$-body Bose and Fermi systems with Quadratic Pair Potentials in one dimension. The originally existed first excited state level is missing in one dimension, which results from the operation of symmetry or antisymmetry of identical particles. In two and higher dimensions, we give all the eigenvalues and the analytical ground state wave functions and the number of degeneracy. Through the comparison with Avinash Khare's results, we have perfected his results.

[ { "created": "Mon, 8 Aug 2011 05:37:08 GMT", "version": "v1" }, { "created": "Tue, 6 Sep 2011 14:03:32 GMT", "version": "v2" }, { "created": "Sun, 9 Oct 2011 01:04:41 GMT", "version": "v3" }, { "created": "Thu, 23 Feb 2012 15:08:48 GMT", "version": "v4" } ]

2015-05-30

[ [ "Wang", "Zhaoliang", "" ], [ "Wang", "Anmin", "" ], [ "Yang", "Yang", "" ], [ "Li", "Xuechao", "" ] ]

Mon, 8 Aug 2011 05:37:08 GMT

2011-08-08T05:37:08Z

[ "physics", "math" ]

arxiv

null

1803.01729

Daniel Lum

Daniel J. Lum, Samuel H. Knarr, and John C. Howell

Frequency-modulated continuous-wave LiDAR compressive depth-mapping

null

Opt. Express 26, 15420-15435 (2018)

10.1364/OE.26.015420

null

eess.SP physics.app-ph physics.optics

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present an inexpensive architecture for converting a frequency-modulated continuous-wave LiDAR system into a compressive-sensing based depth-mapping camera. Instead of raster scanning to obtain depth-maps, compressive sensing is used to significantly reduce the number of measurements. Ideally, our approach requires two difference detectors. % but can operate with only one at the cost of doubling the number of measurments. Due to the large flux entering the detectors, the signal amplification from heterodyne detection, and the effects of background subtraction from compressive sensing, the system can obtain higher signal-to-noise ratios over detector-array based schemes while scanning a scene faster than is possible through raster-scanning. %Moreover, we show how a single total-variation minimization and two fast least-squares minimizations, instead of a single complex nonlinear minimization, can efficiently recover high-resolution depth-maps with minimal computational overhead. Moreover, by efficiently storing only $2m$ data points from $m<n$ measurements of an $n$ pixel scene, we can easily extract depths by solving only two linear equations with efficient convex-optimization methods.

[ { "created": "Fri, 2 Mar 2018 16:13:43 GMT", "version": "v1" }, { "created": "Wed, 16 May 2018 17:29:01 GMT", "version": "v2" } ]

2018-06-06

[ [ "Lum", "Daniel J.", "" ], [ "Knarr", "Samuel H.", "" ], [ "Howell", "John C.", "" ] ]

Fri, 2 Mar 2018 16:13:43 GMT

2018-03-02T16:13:43Z

[ "physics", "eess" ]

arxiv

null

1304.3284

Dmitry Kramkov

Existence and uniqueness of Arrow-Debreu equilibria with consumptions in $\mathbf{L}^0_+$

12 pages; a few typos were corrected and additional explanations were provided; submitted

Theory of Probability and Its Applications, Vol 60, No 4, 2016, pp. 688-695

10.1137/S0040585X97T987958

null

math.PR q-fin.GN q-fin.PM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We consider an economy where agents' consumption sets are given by the cone $\mathbf{L}^0_+$ of non-negative measurable functions and whose preferences are defined by additive utilities satisfying the Inada conditions. We extend to this setting the results in \citet{Dana:93} on the existence and uniqueness of Arrow-Debreu equilibria. In the case of existence, our conditions are necessary and sufficient.

[ { "created": "Thu, 11 Apr 2013 13:16:59 GMT", "version": "v1" }, { "created": "Sun, 12 May 2013 22:19:03 GMT", "version": "v2" } ]

2017-01-31

[ [ "Kramkov", "Dmitry", "" ] ]

Thu, 11 Apr 2013 13:16:59 GMT

2013-04-11T13:16:59Z

[ "math", "q-fin" ]

arxiv

null

1206.3870

David Orden

Lali Barri\`ere, Clemens Huemer, Dieter Mitsche, David Orden

On the Fiedler value of large planar graphs

21 pages, 4 figures, 1 table. Version accepted in Linear Algebra and Its Applications

Linear Algebra and its Applications 439:7 (2013), 2070-2084

10.1016/j.laa.2013.05.032

null

math.CO cs.DM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The Fiedler value $\lambda_2$, also known as algebraic connectivity, is the second smallest Laplacian eigenvalue of a graph. We study the maximum Fiedler value among all planar graphs $G$ with $n$ vertices, denoted by $\lambda_{2\max}$, and we show the bounds $2+\Theta(\frac{1}{n^2}) \leq \lambda_{2\max} \leq 2+O(\frac{1}{n})$. We also provide bounds on the maximum Fiedler value for the following classes of planar graphs: Bipartite planar graphs, bipartite planar graphs with minimum vertex degree~3, and outerplanar graphs. Furthermore, we derive almost tight bounds on $\lambda_{2\max}$ for two more classes of graphs, those of bounded genus and $K_h$-minor-free graphs.

[ { "created": "Mon, 18 Jun 2012 09:51:45 GMT", "version": "v1" }, { "created": "Fri, 7 Jun 2013 07:01:38 GMT", "version": "v2" } ]

2020-07-21

[ [ "Barrière", "Lali", "" ], [ "Huemer", "Clemens", "" ], [ "Mitsche", "Dieter", "" ], [ "Orden", "David", "" ] ]

Mon, 18 Jun 2012 09:51:45 GMT

2012-06-18T09:51:45Z

[ "cs", "math" ]

arxiv

null

2211.01703

Ke Sun

Ke Sun, Samir M. Perlaza, and Alain Jean-Marie

$2 \times 2$ Zero-Sum Games with Commitments and Noisy Observations

Accepted by 2023 IEEE Int. Symp. on Information Theory (ISIT)

null

cs.GT cs.IT cs.LG math.IT math.ST stat.ML stat.TH

http://creativecommons.org/licenses/by-nc-nd/4.0/

In this paper, $2\times2$ zero-sum games are studied under the following assumptions: $(1)$ One of the players (the leader) commits to choose its actions by sampling a given probability measure (strategy); $(2)$ The leader announces its action, which is observed by its opponent (the follower) through a binary channel; and $(3)$ the follower chooses its strategy based on the knowledge of the leader's strategy and the noisy observation of the leader's action. Under these conditions, the equilibrium is shown to always exist. Interestingly, even subject to noise, observing the actions of the leader is shown to be either beneficial or immaterial for the follower. More specifically, the payoff at the equilibrium of this game is upper bounded by the payoff at the Stackelberg equilibrium (SE) in pure strategies; and lower bounded by the payoff at the Nash equilibrium, which is equivalent to the SE in mixed strategies.Finally, necessary and sufficient conditions for observing the payoff at equilibrium to be equal to its lower bound are presented. Sufficient conditions for the payoff at equilibrium to be equal to its upper bound are also presented.

[ { "created": "Thu, 3 Nov 2022 10:56:00 GMT", "version": "v1" }, { "created": "Wed, 10 May 2023 00:50:30 GMT", "version": "v2" }, { "created": "Thu, 11 May 2023 08:29:01 GMT", "version": "v3" } ]

2023-05-12

[ [ "Sun", "Ke", "" ], [ "Perlaza", "Samir M.", "" ], [ "Jean-Marie", "Alain", "" ] ]

Thu, 3 Nov 2022 10:56:00 GMT

2022-11-03T10:56:00Z

[ "cs", "math", "stat" ]

arxiv

null

2004.05839

Simone Garatti

Marco C. Campi and Simone Garatti

A Theory of the Risk for Optimization with Relaxation and its Application to Support Vector Machines

https://www.jmlr.org/papers/v22/21-0641.html

Journal of Machine Learning Research 22(288):1-38, 2021

null

cs.LG cs.SY eess.SY math.OC stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this paper we consider optimization with relaxation, an ample paradigm to make data-driven designs. This approach was previously considered by the same authors of this work in Garatti and Campi (2019), a study that revealed a deep-seated connection between two concepts: risk (probability of not satisfying a new, out-of-sample, constraint) and complexity (according to a definition introduced in paper Garatti and Campi (2019)). This connection was shown to have profound implications in applications because it implied that the risk can be estimated from the complexity, a quantity that can be measured from the data without any knowledge of the data-generation mechanism. In the present work we establish new results. First, we expand the scope of Garatti and Campi (2019) so as to embrace a more general setup that covers various algorithms in machine learning. Then, we study classical support vector methods - including SVM (Support Vector Machine), SVR (Support Vector Regression) and SVDD (Support Vector Data Description) - and derive new results for the ability of these methods to generalize. All results are valid for any finite size of the data set. When the sample size tends to infinity, we establish the unprecedented result that the risk approaches the ratio between the complexity and the cardinality of the data sample, regardless of the value of the complexity.

[ { "created": "Mon, 13 Apr 2020 09:38:25 GMT", "version": "v1" }, { "created": "Wed, 30 Sep 2020 10:34:10 GMT", "version": "v2" }, { "created": "Tue, 20 Oct 2020 19:26:04 GMT", "version": "v3" }, { "created": "Mon, 8 Jan 2024 11:00:50 GMT", "version": "v4" } ]

2024-01-09

[ [ "Campi", "Marco C.", "" ], [ "Garatti", "Simone", "" ] ]

Mon, 13 Apr 2020 09:38:25 GMT

2020-04-13T09:38:25Z

[ "stat", "cs", "eess", "math" ]

arxiv

null

2106.12727

Cuimin Ba

Robust Misspecified Models and Paradigm Shifts

null

econ.TH

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Individuals use models to guide decisions, but many models are wrong. This paper studies which misspecified models are likely to persist when individuals also entertain alternative models. Consider an agent who uses her model to learn the relationship between action choices and outcomes. The agent exhibits sticky model switching, captured by a threshold rule such that she switches to an alternative model when it is a sufficiently better fit for the data she observes. The main result provides a characterization of whether a model persists based on two key features that are straightforward to derive from the primitives of the learning environment, namely, the model's asymptotic accuracy in predicting the equilibrium pattern of observed outcomes and the 'tightness' of the prior around this equilibrium. I show that misspecified models can be robust in that they persist against a wide range of competing models -- including the correct model -- despite individuals observing an infinite amount of data. Moreover, simple misspecified models with entrenched priors can be even more robust than correctly specified models. I use this characterization to provide a learning foundation for the persistence of systemic biases in two applications. First, in an effort-choice problem, I show that overconfidence in one's ability is more robust than underconfidence. Second, a simplistic binary view of politics is more robust than the more complex correct view when individuals consume media without fully recognizing the reporting bias.

[ { "created": "Thu, 24 Jun 2021 02:18:22 GMT", "version": "v1" }, { "created": "Sun, 20 Mar 2022 00:51:23 GMT", "version": "v2" }, { "created": "Mon, 21 Aug 2023 09:56:28 GMT", "version": "v3" } ]

2023-08-22

[ [ "Ba", "Cuimin", "" ] ]

Thu, 24 Jun 2021 02:18:22 GMT

2021-06-24T02:18:22Z

[ "econ" ]

arxiv

null

Distinguishing the Relative Impact of Post-Traumatic Stress Disorder and Traumatic Brain Injury on iPad-Measured Cognitive Function.

null

Traumatic brain injury (TBI) and post-traumatic stress disorder (PTSD) are highly prevalent, and frequently comorbid, among active and retired military service members. Both TBI and PTSD may contribute to impaired cognitive function, but it remains insufficiently clear what the relative impact of each is on overall cognition and whether multiple TBIs may further impair cognitive function. To understand the relative impact of TBI and symptoms of PTSD on cognitive function we examined data from 326 active or retired military service members, or dependents, either with or without a history of TBI, using questionnaires and the NIH Toolbox Cognitive Battery (NIH-TB), a brief iPad-based assessment that measures the cognitive domains most important to daily functioning. The NIH-TB was developed for use as a "common currency" among research studies, and was more recently adapted to the iPad for ease of use. To our knowledge, this is the first report of its application to evaluate the relative impact of TBI and PTSD. Our results indicate that cognitive function remains largely intact after multiple TBIs if symptoms of PTSD are not evident, and that measures of literacy and overall intelligence are relatively impervious to both TBI and PTSD. When cognitive impairment is observed after TBI, it is predominantly associated with the presence of significant symptoms of PTSD in most domains. However, TBI alone may impair some aspects of executive function. These findings need to be validated in other populations.

null

pubmed

null

2,019

null

31,841,649

Cyberpsychology, behavior and social networking

2110.06482

Yifei Wang

Yifei Wang, Tolga Ergen and Mert Pilanci

Parallel Deep Neural Networks Have Zero Duality Gap

null

cs.LG math.OC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Training deep neural networks is a challenging non-convex optimization problem. Recent work has proven that the strong duality holds (which means zero duality gap) for regularized finite-width two-layer ReLU networks and consequently provided an equivalent convex training problem. However, extending this result to deeper networks remains to be an open problem. In this paper, we prove that the duality gap for deeper linear networks with vector outputs is non-zero. In contrast, we show that the zero duality gap can be obtained by stacking standard deep networks in parallel, which we call a parallel architecture, and modifying the regularization. Therefore, we prove the strong duality and existence of equivalent convex problems that enable globally optimal training of deep networks. As a by-product of our analysis, we demonstrate that the weight decay regularization on the network parameters explicitly encourages low-rank solutions via closed-form expressions. In addition, we show that strong duality holds for three-layer standard ReLU networks given rank-1 data matrices.

[ { "created": "Wed, 13 Oct 2021 04:03:51 GMT", "version": "v1" }, { "created": "Thu, 23 Jun 2022 22:50:57 GMT", "version": "v2" }, { "created": "Tue, 7 Mar 2023 03:01:00 GMT", "version": "v3" } ]

2023-03-08

[ [ "Wang", "Yifei", "" ], [ "Ergen", "Tolga", "" ], [ "Pilanci", "Mert", "" ] ]

Wed, 13 Oct 2021 04:03:51 GMT

2021-10-13T04:03:51Z

[ "cs", "math" ]

arxiv

null

1411.0721

Ariel Rokem

Ariel Rokem, Jason D. Yeatman, Franco Pestilli, Kendrick N. Kay, Aviv Mezer, Stefan van der Walt, and Brian A. Wandell

Evaluating the accuracy of diffusion MRI models in white matter

null

10.1371/journal.pone.0123272

null

q-bio.QM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Models of diffusion MRI within a voxel are useful for making inferences about the properties of the tissue and inferring fiber orientation distribution used by tractography algorithms. A useful model must fit the data accurately. However, evaluations of model-accuracy of some of the models that are commonly used in analyzing human white matter have not been published before. Here, we evaluate model-accuracy of the two main classes of diffusion MRI models. The diffusion tensor model (DTM) summarizes diffusion as a 3-dimensional Gaussian distribution. Sparse fascicle models (SFM) summarize the signal as a linear sum of signals originating from a collection of fascicles oriented in different directions. We use cross-validation to assess model-accuracy at different gradient amplitudes (b-values) throughout the white matter. Specifically, we fit each model to all the white matter voxels in one data set and then use the model to predict a second, independent data set. This is the first evaluation of model-accuracy of these models. In most of the white matter the DTM predicts the data more accurately than test-retest reliability; SFM model-accuracy is higher than test-retest reliability and also higher than the DTM, particularly for measurements with (a) a b-value above 1000 in locations containing fiber crossings, and (b) in the regions of the brain surrounding the optic radiations. The SFM also has better parameter-validity: it more accurately estimates the fiber orientation distribution function (fODF) in each voxel, which is useful for fiber tracking.

[ { "created": "Mon, 3 Nov 2014 22:25:33 GMT", "version": "v1" }, { "created": "Mon, 10 Nov 2014 20:58:41 GMT", "version": "v2" }, { "created": "Sat, 14 Mar 2015 20:00:29 GMT", "version": "v3" } ]

2017-02-08

[ [ "Rokem", "Ariel", "" ], [ "Yeatman", "Jason D.", "" ], [ "Pestilli", "Franco", "" ], [ "Kay", "Kendrick N.", "" ], [ "Mezer", "Aviv", "" ], [ "van der Walt", "Stefan", "" ], [ "Wandell", "Brian A.", "" ] ]

Mon, 3 Nov 2014 22:25:33 GMT

2014-11-03T22:25:33Z

[ "q-bio" ]

arxiv

null

2007.06075

Jo\~ao M Pereira

Ali Hasan, Jo\~ao M. Pereira, Sina Farsiu, Vahid Tarokh

Identifying Latent Stochastic Differential Equations

20 pages, 8 figures, to be published in IEEE Transactions of Signal Processing

null

stat.ML cs.LG

http://creativecommons.org/licenses/by/4.0/

We present a method for learning latent stochastic differential equations (SDEs) from high-dimensional time series data. Given a high-dimensional time series generated from a lower dimensional latent unknown It\^o process, the proposed method learns the mapping from ambient to latent space, and the underlying SDE coefficients, through a self-supervised learning approach. Using the framework of variational autoencoders, we consider a conditional generative model for the data based on the Euler-Maruyama approximation of SDE solutions. Furthermore, we use recent results on identifiability of latent variable models to show that the proposed model can recover not only the underlying SDE coefficients, but also the original latent variables, up to an isometry, in the limit of infinite data. We validate the method through several simulated video processing tasks, where the underlying SDE is known, and through real world datasets.

[ { "created": "Sun, 12 Jul 2020 19:46:31 GMT", "version": "v1" }, { "created": "Sun, 11 Oct 2020 14:00:31 GMT", "version": "v2" }, { "created": "Wed, 14 Oct 2020 16:28:17 GMT", "version": "v3" }, { "created": "Thu, 6 May 2021 14:55:48 GMT", "version": "v4" }, { "created": "Fri, 26 Nov 2021 23:41:49 GMT", "version": "v5" } ]

2021-11-30

[ [ "Hasan", "Ali", "" ], [ "Pereira", "João M.", "" ], [ "Farsiu", "Sina", "" ], [ "Tarokh", "Vahid", "" ] ]

Sun, 12 Jul 2020 19:46:31 GMT

2020-07-12T19:46:31Z

[ "cs", "stat" ]

arxiv

null

1804.05534

\'Alvaro L\'opez-Ravent\'os

\'Alvaro L\'opez-Ravent\'os, Francesc Wilhelmi, Sergio Barrachina-Mu\~noz and Boris Bellalta

Combining Software Defined Networks and Machine Learning to enable Self Organizing WLANs

null

cs.NI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Next generation of wireless local area networks (WLANs) will operate in dense, chaotic and highly dynamic scenarios that in a significant number of cases may result in a low user experience due to uncontrolled high interference levels. Flexible network architectures, such as the software-defined networking (SDN) paradigm, will provide WLANs with new capabilities to deal with users' demands, while achieving greater levels of efficiency and flexibility in those complex scenarios. On top of SDN, the use of machine learning (ML) techniques may improve network resource usage and management by identifying feasible configurations through learning. ML techniques can drive WLANs to reach optimal working points by means of parameter adjustment, in order to cope with different network requirements and policies, as well as with the dynamic conditions. In this paper we overview the work done in SDN for WLANs, as well as the pioneering works considering ML for WLAN optimization. Finally, in order to demonstrate the potential of ML techniques in combination with SDN to improve the network operation, we evaluate different use cases for intelligent-based spatial reuse and dynamic channel bonding operation in WLANs using Multi-Armed Bandits.

[ { "created": "Mon, 16 Apr 2018 07:53:28 GMT", "version": "v1" }, { "created": "Thu, 1 Aug 2019 14:16:37 GMT", "version": "v2" }, { "created": "Thu, 12 Sep 2019 13:04:01 GMT", "version": "v3" } ]

2019-09-13

[ [ "López-Raventós", "Álvaro", "" ], [ "Wilhelmi", "Francesc", "" ], [ "Barrachina-Muñoz", "Sergio", "" ], [ "Bellalta", "Boris", "" ] ]

Mon, 16 Apr 2018 07:53:28 GMT

2018-04-16T07:53:28Z

[ "cs" ]

arxiv

null

0904.2385

Stefan Milius

Stefan Milius, Lawrence S. Moss

The Category Theoretic Solution of Recursive Program Schemes

this version includes the corrections from the corrigendum in Theoret. Comput. Sci. 403 (2008), 409-415

Theoret. Comput. Sci. 366 (2006), 3-59

null

cs.LO math.CT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper provides a general account of the notion of recursive program schemes, studying both uninterpreted and interpreted solutions. It can be regarded as the category-theoretic version of the classical area of algebraic semantics. The overall assumptions needed are small indeed: working only in categories with "enough final coalgebras" we show how to formulate, solve, and study recursive program schemes. Our general theory is algebraic and so avoids using ordered, or metric structures. Our work generalizes the previous approaches which do use this extra structure by isolating the key concepts needed to study substitution in infinite trees, including second-order substitution. As special cases of our interpreted solutions we obtain the usual denotational semantics using complete partial orders, and the one using complete metric spaces. Our theory also encompasses implicitly defined objects which are not usually taken to be related to recursive program schemes. For example, the classical Cantor two-thirds set falls out as an interpreted solution (in our sense) of a recursive program scheme.

[ { "created": "Wed, 15 Apr 2009 19:57:26 GMT", "version": "v1" }, { "created": "Wed, 15 Apr 2009 20:27:58 GMT", "version": "v2" } ]

2011-01-26

[ [ "Milius", "Stefan", "" ], [ "Moss", "Lawrence S.", "" ] ]

Wed, 15 Apr 2009 19:57:26 GMT

2009-04-15T19:57:26Z

[ "cs", "math" ]

arxiv

null

1811.06586

Matteo Cococcioni

Matteo Cococcioni and Nicola Marzari

Energetics and cathode voltages of LiMPO$_4$ olivines (M = Fe, Mn) from extended Hubbard functionals

null

Phys. Rev. Materials 3, 033801 (2019)

10.1103/PhysRevMaterials.3.033801

null

cond-mat.mtrl-sci

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Transition-metal compounds pose serious challenges to first-principles calculations based on density-functional theory (DFT), due to the inability of most approximate exchange-correlation functionals to capture the localization of valence electrons on their $d$ states, essential for a predictive modeling of their properties. In this work we focus on two representatives of a well known family of cathode materials for Li-ion batteries, namely the orthorhombic LiMPO$_4$ olivines (M = Fe, Mn). We show that extended Hubbard functionals with on-site ($U$) and inter-site ($V$) interactions (so called DFT+U+V) can predict the electronic structure of the mixed-valence phases, the formation energy of the materials with intermediate Li contents, and the overall average voltage of the battery with remarkable accuracy. We find, in particular, that the inclusion of inter-site interactions in the corrective Hamiltonian improves considerably the prediction of thermodynamic quantities when electronic localization occurs in the presence of significant interatomic hybridization (as is the case for the Mn compound), and that the self-consistent evaluation of the effective interaction parameters as material- and ground-state-dependent quantities allows the prediction of energy differences between different phases and concentrations.

[ { "created": "Thu, 15 Nov 2018 20:47:48 GMT", "version": "v1" }, { "created": "Thu, 21 Feb 2019 21:48:01 GMT", "version": "v2" }, { "created": "Tue, 26 Feb 2019 11:35:28 GMT", "version": "v3" } ]

2019-03-13

[ [ "Cococcioni", "Matteo", "" ], [ "Marzari", "Nicola", "" ] ]

Thu, 15 Nov 2018 20:47:48 GMT

2018-11-15T20:47:48Z

[ "physics" ]

arxiv

null

1810.12862

David Hong

David Hong and Fan Yang and Jeffrey A. Fessler and Laura Balzano

Optimally Weighted PCA for High-Dimensional Heteroscedastic Data

39 pages, 9 figures

null

math.ST stat.TH

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Modern data are increasingly both high-dimensional and heteroscedastic. This paper considers the challenge of estimating underlying principal components from high-dimensional data with noise that is heteroscedastic across samples, i.e., some samples are noisier than others. Such heteroscedasticity naturally arises, e.g., when combining data from diverse sources or sensors. A natural way to account for this heteroscedasticity is to give noisier blocks of samples less weight in PCA by using the leading eigenvectors of a weighted sample covariance matrix. We consider the problem of choosing weights to optimally recover the underlying components. In general, one cannot know these optimal weights since they depend on the underlying components we seek to estimate. However, we show that under some natural statistical assumptions the optimal weights converge to a simple function of the signal and noise variances for high-dimensional data. Surprisingly, the optimal weights are not the inverse noise variance weights commonly used in practice. We demonstrate the theoretical results through numerical simulations and comparisons with existing weighting schemes. Finally, we briefly discuss how estimated signal and noise variances can be used when the true variances are unknown, and we illustrate the optimal weights on real data from astronomy.

[ { "created": "Tue, 30 Oct 2018 17:06:25 GMT", "version": "v1" }, { "created": "Wed, 21 Nov 2018 22:51:21 GMT", "version": "v2" }, { "created": "Sat, 8 Jan 2022 20:48:21 GMT", "version": "v3" }, { "created": "Tue, 13 Sep 2022 11:49:56 GMT", "version": "v4" } ]

2022-09-14

[ [ "Hong", "David", "" ], [ "Yang", "Fan", "" ], [ "Fessler", "Jeffrey A.", "" ], [ "Balzano", "Laura", "" ] ]

Tue, 30 Oct 2018 17:06:25 GMT

2018-10-30T17:06:25Z

[ "math", "stat" ]

arxiv

null

2112.06649

Maxim Ziatdinov

Maxim Ziatdinov, Yongtao Liu, Anna N. Morozovska, Eugene A. Eliseev, Xiaohang Zhang, Ichiro Takeuchi, and Sergei V. Kalinin

Hypothesis Learning in Automated Experiment: Application to Combinatorial Materials Libraries

Fixed typo in Eq. 1. Expanded the introduction part. The code reproducing Algorithm 1 is available at https://github.com/ziatdinovmax/hypoAL

Adv. Mater. 2022, 2201345

10.1002/adma.202201345

null

cond-mat.mtrl-sci physics.data-an

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Machine learning is rapidly becoming an integral part of experimental physical discovery via automated and high-throughput synthesis, and active experiments in scattering and electron/probe microscopy. This, in turn, necessitates the development of active learning methods capable of exploring relevant parameter spaces with the smallest number of steps. Here we introduce an active learning approach based on co-navigation of the hypothesis and experimental spaces. This is realized by combining the structured Gaussian Processes containing probabilistic models of the possible system's behaviors (hypotheses) with reinforcement learning policy refinement (discovery). This approach closely resembles classical human-driven physical discovery, when several alternative hypotheses realized via models with adjustable parameters are tested during an experiment. We demonstrate this approach for exploring concentration-induced phase transitions in combinatorial libraries of Sm-doped BiFeO3 using Piezoresponse Force Microscopy, but it is straightforward to extend it to higher-dimensional parameter spaces and more complex physical problems once the experimental workflow and hypothesis-generation are available.

[ { "created": "Mon, 13 Dec 2021 18:03:15 GMT", "version": "v1" }, { "created": "Thu, 10 Mar 2022 01:15:13 GMT", "version": "v2" }, { "created": "Thu, 21 Apr 2022 03:11:16 GMT", "version": "v3" } ]

2022-04-22

[ [ "Ziatdinov", "Maxim", "" ], [ "Liu", "Yongtao", "" ], [ "Morozovska", "Anna N.", "" ], [ "Eliseev", "Eugene A.", "" ], [ "Zhang", "Xiaohang", "" ], [ "Takeuchi", "Ichiro", "" ], [ "Kalinin", "Sergei V.", "" ] ]

Mon, 13 Dec 2021 18:03:15 GMT

2021-12-13T18:03:15Z

[ "physics" ]

arxiv

null

1811.10344

James Gray

James Richard Andrew Gray

Hall's criterion for nilpotence in semi-abelian categories

null

Advances in Mathematics 349 (2019) 911-919

10.1016/j.aim.2019.04.025.

null

math.CT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

A well-known theorem of P. Hall, usually called Hall's criterion for nilpotence, states: a group G is nilpotent whenever it has a normal subgroup N such that G/[N,N] and N are nilpotent. We widely generalize this result, replacing groups with objects in an abstract semi-abelian category satisfying suitable conditions. In particular, these conditions are satisfied in any algebraically coherent semi-abelian category, and hence in many categories of classical algebraic structures (including a few where Hall's theorem is already known). Note that the categories of crossed modules, crossed Lie algebras, n-cat groups, and cocommutative Hopf algebras over fields are also algebraically coherent. We give, however, a counter-example showing that Hall's theorem does not hold in the (semi-abelian) variety of non-associative rings.

[ { "created": "Mon, 26 Nov 2018 13:03:28 GMT", "version": "v1" }, { "created": "Fri, 22 Mar 2019 10:44:03 GMT", "version": "v2" }, { "created": "Mon, 8 Apr 2019 07:26:33 GMT", "version": "v3" }, { "created": "Sat, 4 May 2019 04:19:24 GMT", "version": "v4" } ]

2019-05-07

[ [ "Gray", "James Richard Andrew", "" ] ]

Mon, 26 Nov 2018 13:03:28 GMT

2018-11-26T13:03:28Z

[ "math" ]

arxiv

null

1705.08562

Kun He

Kun He, Fatih Cakir, Sarah Adel Bargal, Stan Sclaroff

Hashing as Tie-Aware Learning to Rank

15 pages, 3 figures. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018

null

stat.ML cs.CV cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Hashing, or learning binary embeddings of data, is frequently used in nearest neighbor retrieval. In this paper, we develop learning to rank formulations for hashing, aimed at directly optimizing ranking-based evaluation metrics such as Average Precision (AP) and Normalized Discounted Cumulative Gain (NDCG). We first observe that the integer-valued Hamming distance often leads to tied rankings, and propose to use tie-aware versions of AP and NDCG to evaluate hashing for retrieval. Then, to optimize tie-aware ranking metrics, we derive their continuous relaxations, and perform gradient-based optimization with deep neural networks. Our results establish the new state-of-the-art for image retrieval by Hamming ranking in common benchmarks.

[ { "created": "Tue, 23 May 2017 23:42:46 GMT", "version": "v1" }, { "created": "Thu, 23 Nov 2017 02:07:11 GMT", "version": "v2" }, { "created": "Wed, 28 Mar 2018 04:42:56 GMT", "version": "v3" }, { "created": "Tue, 9 Oct 2018 20:37:00 GMT", "version": "v4" } ]

2018-10-11

[ [ "He", "Kun", "" ], [ "Cakir", "Fatih", "" ], [ "Bargal", "Sarah Adel", "" ], [ "Sclaroff", "Stan", "" ] ]

Tue, 23 May 2017 23:42:46 GMT

2017-05-23T23:42:46Z

[ "cs", "stat" ]

arxiv

null

1811.00207

Eric Smith

Hannah Rashkin, Eric Michael Smith, Margaret Li and Y-Lan Boureau

Towards Empathetic Open-domain Conversation Models: a New Benchmark and Dataset

accepted to ACL 2019 (long paper)

null

cs.CL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

One challenge for dialogue agents is recognizing feelings in the conversation partner and replying accordingly, a key communicative skill. While it is straightforward for humans to recognize and acknowledge others' feelings in a conversation, this is a significant challenge for AI systems due to the paucity of suitable publicly-available datasets for training and evaluation. This work proposes a new benchmark for empathetic dialogue generation and EmpatheticDialogues, a novel dataset of 25k conversations grounded in emotional situations. Our experiments indicate that dialogue models that use our dataset are perceived to be more empathetic by human evaluators, compared to models merely trained on large-scale Internet conversation data. We also present empirical comparisons of dialogue model adaptations for empathetic responding, leveraging existing models or datasets without requiring lengthy re-training of the full model.

[ { "created": "Thu, 1 Nov 2018 03:43:10 GMT", "version": "v1" }, { "created": "Mon, 4 Feb 2019 18:33:17 GMT", "version": "v2" }, { "created": "Wed, 19 Jun 2019 04:09:22 GMT", "version": "v3" }, { "created": "Fri, 26 Jul 2019 18:24:13 GMT", "version": "v4" }, { "created": "Wed, 28 Aug 2019 21:35:38 GMT", "version": "v5" } ]

2019-08-30

[ [ "Rashkin", "Hannah", "" ], [ "Smith", "Eric Michael", "" ], [ "Li", "Margaret", "" ], [ "Boureau", "Y-Lan", "" ] ]

Thu, 1 Nov 2018 03:43:10 GMT

2018-11-01T03:43:10Z

[ "cs" ]

arxiv

null

2003.02824

Min-Hung Chen

Min-Hung Chen, Baopu Li, Yingze Bao, Ghassan AlRegib, Zsolt Kira

Action Segmentation with Joint Self-Supervised Temporal Domain Adaptation

CVPR 2020. Source code: https://github.com/cmhungsteve/SSTDA

null

cs.CV cs.LG eess.IV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Despite the recent progress of fully-supervised action segmentation techniques, the performance is still not fully satisfactory. One main challenge is the problem of spatiotemporal variations (e.g. different people may perform the same activity in various ways). Therefore, we exploit unlabeled videos to address this problem by reformulating the action segmentation task as a cross-domain problem with domain discrepancy caused by spatio-temporal variations. To reduce the discrepancy, we propose Self-Supervised Temporal Domain Adaptation (SSTDA), which contains two self-supervised auxiliary tasks (binary and sequential domain prediction) to jointly align cross-domain feature spaces embedded with local and global temporal dynamics, achieving better performance than other Domain Adaptation (DA) approaches. On three challenging benchmark datasets (GTEA, 50Salads, and Breakfast), SSTDA outperforms the current state-of-the-art method by large margins (e.g. for the F1@25 score, from 59.6% to 69.1% on Breakfast, from 73.4% to 81.5% on 50Salads, and from 83.6% to 89.1% on GTEA), and requires only 65% of the labeled training data for comparable performance, demonstrating the usefulness of adapting to unlabeled target videos across variations. The source code is available at https://github.com/cmhungsteve/SSTDA.

[ { "created": "Thu, 5 Mar 2020 18:52:33 GMT", "version": "v1" }, { "created": "Fri, 6 Mar 2020 17:16:37 GMT", "version": "v2" }, { "created": "Wed, 18 Mar 2020 21:09:07 GMT", "version": "v3" } ]

2020-03-20

[ [ "Chen", "Min-Hung", "" ], [ "Li", "Baopu", "" ], [ "Bao", "Yingze", "" ], [ "AlRegib", "Ghassan", "" ], [ "Kira", "Zsolt", "" ] ]

Thu, 5 Mar 2020 18:52:33 GMT

2020-03-05T18:52:33Z

[ "cs", "eess" ]

arxiv

null

1505.06028

Sergey Balashev

S.A. Balashev, E.E. Kholupenko, J. Chluba, A.V. Ivanchik, D.A. Varshalovich

Spectral distortions of the CMB dipole

6 pages, 2 figures, published in ApJ

ApJ, 810, 131, 2015

10.1088/0004-637X/810/2/131

null

astro-ph.CO

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We consider the distortions of the cosmic microwave background (CMB) dipole anisotropy related to the primordial recombination radiation (PRR) and primordial $y$- and $\mu$-distortions. The signals arise due to our motion relative to the CMB restframe and appear as a frequency-dependent distortion of the CMB temperature dipole. To leading order, the expected relative distortion of the CMB dipole does not depend on the particular observation directions and reaches the level of $10^{-6}$ for the PRR- and $\mu$-distortions and $10^{-5}$ for the $y$-distortion in the frequency range 1 -- 700 GHz. The temperature differences arising from the dipole anisotropy of the relic CMB distortions depend on the observation directions. For mutually opposite directions, collinear to the CMB dipole axis, the temperature differences because of the PRR- and $\mu$-dipole anisotropy attain values $\Delta T\simeq 10\,$nK in the considered range. The temperature difference arising from the $y$-dipole anisotropy may reach values of up to $1\,\mu$K. The key features of the considered effect are as follow: (i) an observation of the effect does not require absolute calibration; (ii) patches of sky with minimal foreground contamination can be chosen. Future measurements of the CMB dipole distortion thus will provide an alternative method for direct detection of the PRR-, $y$-, and $\mu$-distortions. The $y$-distortion dipole may be detectable with PIXIE at a few standard deviations.

[ { "created": "Fri, 22 May 2015 11:11:07 GMT", "version": "v1" }, { "created": "Fri, 11 Sep 2015 11:17:14 GMT", "version": "v2" } ]

2015-09-14

[ [ "Balashev", "S. A.", "" ], [ "Kholupenko", "E. E.", "" ], [ "Chluba", "J.", "" ], [ "Ivanchik", "A. V.", "" ], [ "Varshalovich", "D. A.", "" ] ]

Fri, 22 May 2015 11:11:07 GMT

2015-05-22T11:11:07Z

[ "physics" ]

arxiv

null

Modeling Temporality of Human Intentions by Domain Adaptation

null

Categorizing patient`s intentions in conversational assessment can help decision making in clinical treatments. Many conversation corpora span broaden a series of time stages. However, it is not clear that how the themes shift in the conversation impact on the performance of human intention categorization (eg., patients might show different behaviors during the beginning versus the end). This paper proposes a method that models the temporal factor by using domain adaptation on clinical dialogue corpora, Motivational Interviewing (MI). We deploy Bi-LSTM and topic model jointly to learn language usage change across different time sessions. We conduct experiments on the MI corpora to show the promising improvement after considering temporality in the classification task.

null

acl

inproceedings

huang-etal-2018-modeling

Riloff, Ellen and Chiang, David and Hockenmaier, Julia and Tsujii, Jun{'}ichi

oct # "-" # nov

2,018

Brussels, Belgium

Association for Computational Linguistics

https://aclanthology.org/D18-1074/

Huang, Xiaolei and Liu, Lixing and Carey, Kate and Woolley, Joshua and Scherer, Stefan and Borsari, Brian

Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing

696--701

null

1912.07418

Huajun Wang

Huajun Wang, Yuanhai Shao, Shenglong Zhou, Ce Zhang and Naihua Xiu

Support Vector Machine Classifier via $L_{0/1}$ Soft-Margin Loss

null

math.OC cs.LG stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Support vector machine (SVM) has attracted great attentions for the last two decades due to its extensive applications, and thus numerous optimization models have been proposed. To distinguish all of them, in this paper, we introduce a new model equipped with an $L_{0/1}$ soft-margin loss (dubbed as $L_{0/1}$-SVM) which well captures the nature of the binary classification. Many of the existing convex/non-convex soft-margin losses can be viewed as a surrogate of the $L_{0/1}$ soft-margin loss. Despite the discrete nature of $L_{0/1}$, we manage to establish the existence of global minimizer of the new model as well as revealing the relationship among its minimizers and KKT/P-stationary points. These theoretical properties allow us to take advantage of the alternating direction method of multipliers. In addition, the $L_{0/1}$-support vector operator is introduced as a filter to prevent outliers from being support vectors during the training process. Hence, the method is expected to be relatively robust. Finally, numerical experiments demonstrate that our proposed method generates better performance in terms of much shorter computational time with much fewer number of support vectors when against with some other leading methods in areas of SVM. When the data size gets bigger, its advantage becomes more evident.

[ { "created": "Mon, 16 Dec 2019 14:42:30 GMT", "version": "v1" }, { "created": "Fri, 10 Jul 2020 13:36:15 GMT", "version": "v2" }, { "created": "Mon, 15 Mar 2021 03:30:28 GMT", "version": "v3" }, { "created": "Mon, 5 Apr 2021 14:30:46 GMT", "version": "v4" } ]

2021-04-06

[ [ "Wang", "Huajun", "" ], [ "Shao", "Yuanhai", "" ], [ "Zhou", "Shenglong", "" ], [ "Zhang", "Ce", "" ], [ "Xiu", "Naihua", "" ] ]

Mon, 16 Dec 2019 14:42:30 GMT

2019-12-16T14:42:30Z

[ "cs", "math", "stat" ]

arxiv

null

Simon @ {LT}-{EDI}-{EACL}2021: Detecting Hope Speech with {BERT}

null

In today`s society, the rapid development of communication technology allows us to communicate with people from different parts of the world. In the process of communication, each person treats others differently. Some people are used to using offensive and sarcastic language to express their views. These words cause pain to others and make people feel down. Some people are used to sharing happiness with others and encouraging others. Such people bring joy and hope to others through their words. On social media platforms, these two kinds of language are all over the place. If people want to make the online world a better place, they will have to deal with both. So identifying offensive language and hope language is an essential task. There have been many assignments about offensive language. Shared Task on Hope Speech Detection for Equality, Diversity, and Inclusion at LT-EDI 2021-EACL 2021 uses another unique perspective {--} to identify the language of Hope to make contributions to society. The XLM-Roberta model is an excellent multilingual model. Our team used a fine-tuned XLM-Roberta model to accomplish this task.

null

acl

inproceedings

que-2021-simon-lt

Chakravarthi, Bharathi Raja and McCrae, John P. and Zarrouk, Manel and Bali, Kalika and Buitelaar, Paul

apr

2,021

Kyiv

Association for Computational Linguistics

https://aclanthology.org/2021.ltedi-1.26/

Que, Qinyu

Proceedings of the First Workshop on Language Technology for Equality, Diversity and Inclusion

175--179

null

Unsupervised Neural Text Simplification

null

The paper presents a first attempt towards unsupervised neural text simplification that relies only on unlabeled text corpora. The core framework is composed of a shared encoder and a pair of attentional-decoders, crucially assisted by discrimination-based losses and denoising. The framework is trained using unlabeled text collected from en-Wikipedia dump. Our analysis (both quantitative and qualitative involving human evaluators) on public test data shows that the proposed model can perform text-simplification at both lexical and syntactic levels, competitive to existing supervised methods. It also outperforms viable unsupervised baselines. Adding a few labeled pairs helps improve the performance further.

null

acl

inproceedings

surya-etal-2019-unsupervised

Korhonen, Anna and Traum, David and M{\`a}rquez, Llu{\'i}s

jul

2,019

Florence, Italy

Association for Computational Linguistics

https://aclanthology.org/P19-1198/

Surya, Sai and Mishra, Abhijit and Laha, Anirban and Jain, Parag and Sankaranarayanan, Karthik

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics

2058--2068

null

1804.01028

Alex Tourigny-Plante

Alex Tourigny-Plante, Vincent Michaud-Belleau, Nicolas Bourbeau-H\'ebert, Hugo Bergeron, J\'er\^ome Genest and Jean-Daniel Desch\^enes

An open and flexible digital phase-locked loop for optical metrology

This article may be downloaded for personal use only. Any other use requires prior permission of the author and AIP Publishing. This article appeared in Review of Scientific Instruments 89, 093103 (2018) and may be found at https://doi.org/10.1063/1.5039344

Review of Scientific Instruments 89, 093103 (2018)

10.1063/1.5039344

null

eess.SP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper presents an open and flexible digital phase lock loop optimized for laser stabilization systems. It is implemented on a cheap and easily accessible FPGA-based digital electronics platform (Red Pitaya) running a customizable open-source firmware. A PC-based software interface allows controlling the platform and optimizing the loop parameters remotely. Several tools are included to allow measurement of quantities of interest smoothly and rapidly. To demonstrate the platform's capabilities, we built a fiber noise canceler over a $400$~m fiber link. Noise cancellation was achieved over a $30$~kHz bandwidth, a value limited mainly by the delays introduced by the actuator and by the round-trip propagation over the fiber link. We measured a total latency of $565$~ns for the platform itself, limiting the theoretically achievable control bandwidth to approximately 225 kHz.

[ { "created": "Tue, 3 Apr 2018 15:16:43 GMT", "version": "v1" }, { "created": "Mon, 8 Jul 2019 14:14:35 GMT", "version": "v2" } ]

2019-07-09

[ [ "Tourigny-Plante", "Alex", "" ], [ "Michaud-Belleau", "Vincent", "" ], [ "Bourbeau-Hébert", "Nicolas", "" ], [ "Bergeron", "Hugo", "" ], [ "Genest", "Jérôme", "" ], [ "Deschênes", "Jean-Daniel", "" ] ]

Tue, 3 Apr 2018 15:16:43 GMT

2018-04-03T15:16:43Z

[ "eess" ]

arxiv

null

2203.14699

Afroza Shirin

Afroza Shirin, Edl Schamiloglu, and Rafael Fierro

Towards Stable Interstellar Flight: Levitation of a Laser-Propelled Sailcraft*

A computational error have been found. It will be resubmitted when it will be fixed

2022 American Control Conference

10.23919/ACC53348.2022.9867520

null

eess.SY cs.SY

http://creativecommons.org/publicdomain/zero/1.0/

Exploring and traveling to distant stars has long fascinated humanity but has been limited due to the vast distances. The Breakthrough Starshot Program aims at eliminating this limitation by traveling to Alpha Centauri, which is 4.37 light-years away. Thus, it is only possible if a vehicle travels at a substantial fraction of the speed of light. The Breakthrough Starshot Program initiatives to develop a proof-of-concept that is accelerating a sail to relativistic speeds using a laser beam aimed at the sail. At this high speed, while a stable beam riding is one of the crucial concerns of this concept, the dynamic stability analysis of a sail is hardly present in the previous literature. Furthermore, it is important to investigate the dynamic stability in the experiment before driving the sail to relativistic speeds. As a proof of concept, we study the dynamic stability of the sail levitated at a certain height by a laser beam. The sail's dynamics are modeled as a rigid body whose shape is parameterized by a sweep function. We estimate the region of attraction (ROA) for dynamic stability analysis using Lyapunov theory and Sum-of-square (SOS) programming. The ROA confirms how much a levitated sail can tolerate the transverse and angular perturbations. We also conclude on some of the important parameters of the sail that affects the dynamic stability. Simulation results validate our theoretical analysis.

[ { "created": "Thu, 24 Mar 2022 17:37:25 GMT", "version": "v1" }, { "created": "Thu, 31 Mar 2022 04:09:55 GMT", "version": "v2" }, { "created": "Sat, 17 Sep 2022 05:23:41 GMT", "version": "v3" } ]

2022-09-20

[ [ "Shirin", "Afroza", "" ], [ "Schamiloglu", "Edl", "" ], [ "Fierro", "Rafael", "" ] ]

Thu, 24 Mar 2022 17:37:25 GMT

2022-03-24T17:37:25Z

[ "cs", "eess" ]

arxiv

null

2208.10230

Yaosen Min

Yaosen Min, Ye Wei, Peizhuo Wang, Xiaoting Wang, Han Li, Nian Wu, Stefan Bauer, Shuxin Zheng, Yu Shi, Yingheng Wang, Ji Wu, Dan Zhao and Jianyang Zeng

From Static to Dynamic Structures: Improving Binding Affinity Prediction with Graph-Based Deep Learning

Update the content according to the published version on Advanced Science (https://doi.org/10.1002/advs.202405404)

null

10.1002/advs.202405404

null

q-bio.BM cs.LG physics.chem-ph q-bio.QM

http://creativecommons.org/licenses/by-sa/4.0/

Accurate prediction of protein-ligand binding affinities is an essential challenge in structure-based drug design. Despite recent advances in data-driven methods for affinity prediction, their accuracy is still limited, partially because they only take advantage of static crystal structures while the actual binding affinities are generally determined by the thermodynamic ensembles between proteins and ligands. One effective way to approximate such a thermodynamic ensemble is to use molecular dynamics (MD) simulation. Here, an MD dataset containing 3,218 different protein-ligand complexes is curated, and Dynaformer, a graph-based deep learning model is further developed to predict the binding affinities by learning the geometric characteristics of the protein-ligand interactions from the MD trajectories. In silico experiments demonstrated that the model exhibits state-of-the-art scoring and ranking power on the CASF-2016 benchmark dataset, outperforming the methods hitherto reported. Moreover, in a virtual screening on heat shock protein 90 (HSP90) using Dynaformer, 20 candidates are identified and their binding affinities are further experimentally validated. Dynaformer displayed promising results in virtual drug screening, revealing 12 hit compounds (two are in the submicromolar range), including several novel scaffolds. Overall, these results demonstrated that the approach offer a promising avenue for accelerating the early drug discovery process.

[ { "created": "Fri, 19 Aug 2022 14:55:12 GMT", "version": "v1" }, { "created": "Wed, 21 Sep 2022 06:16:46 GMT", "version": "v2" }, { "created": "Sat, 3 Jun 2023 09:58:11 GMT", "version": "v3" }, { "created": "Mon, 2 Sep 2024 07:10:37 GMT", "version": "v4" } ]

2024-09-04

[ [ "Min", "Yaosen", "" ], [ "Wei", "Ye", "" ], [ "Wang", "Peizhuo", "" ], [ "Wang", "Xiaoting", "" ], [ "Li", "Han", "" ], [ "Wu", "Nian", "" ], [ "Bauer", "Stefan", "" ], [ "Zheng", "Shuxin", "" ], [ "Shi", "Yu", "" ], [ "Wang", "Yingheng", "" ], [ "Wu", "Ji", "" ], [ "Zhao", "Dan", "" ], [ "Zeng", "Jianyang", "" ] ]

Fri, 19 Aug 2022 14:55:12 GMT

2022-08-19T14:55:12Z

[ "physics", "q-bio", "cs" ]

arxiv

null

1710.02141

Qilian Yu

Qilian Yu, Hang Li, Yun Liao, Shuguang Cui

Fast Budgeted Influence Maximization over Multi-Action Event Logs

null

cs.SI cs.DS

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In a social network, influence maximization is the problem of identifying a set of users that own the maximum {\it influence ability} across the network. In this paper, a novel credit distribution (CD) based model, termed as the multi-action CD (mCD) model, is introduced to quantify the influence ability of each user, which works with practical datasets where one type of action could be recorded for multiple times. Based on this model, influence maximization is formulated as a submodular maximization problem under a general knapsack constraint, which is NP-hard. An efficient streaming algorithm with one-round scan over the user set is developed to find a suboptimal solution. Specifically, we first solve a special case of knapsack constraints, i.e., a cardinality constraint, and show that the developed streaming algorithm can achieve ($\frac{1}{2}-\epsilon$)-approximation of the optimality. Furthermore, for the general knapsack case, we show that the modified streaming algorithm can achieve ($\frac{1}{3}-\epsilon$)-approximation of the optimality. Finally, experiments are conducted over real Twitter dataset and demonstrate that the mCD model enjoys high accuracy compared to the conventional CD model in estimating the total number of people who get influenced in a social network. Moreover, through the comparison to the conventional CD, non-CD models, and the mCD model with the greedy algorithm on the performance of the influence maximization problem, we show the effectiveness and efficiency of the proposed mCD model with the streaming algorithm.

[ { "created": "Thu, 5 Oct 2017 05:59:48 GMT", "version": "v1" }, { "created": "Tue, 10 Oct 2017 21:56:35 GMT", "version": "v2" }, { "created": "Fri, 26 Jan 2018 06:58:18 GMT", "version": "v3" } ]

2018-01-29

[ [ "Yu", "Qilian", "" ], [ "Li", "Hang", "" ], [ "Liao", "Yun", "" ], [ "Cui", "Shuguang", "" ] ]

Thu, 5 Oct 2017 05:59:48 GMT

2017-10-05T05:59:48Z

[ "cs" ]

arxiv

null

Alteration of the proteome profile of the pancreas in diabetic rats induced by streptozotocin.

null

Type 1 diabetes mellitus (T1DM) is receiving increased attention. To obtain a better understanding of the mechanism underlying T1DM, we performed a proteomic study on a rat model induced by streptozotocin. Pancreatic proteins were separated by two-dimensional gel electrophoresis. Eighteen protein spots were differentially expressed (P<0.05) with 2-fold or more increased or decreased intensity in the diabetic rats as compared with controls, of which 11 protein spots were up-regulated and 7 protein spots were down-regulated. These protein spots were successfully identified by liquid chromatography-electrospray ionization tandem mass spectrometry. The 60 kDa heat shock protein, the carbonyl reductase 1 (Cbr1), the hydroxyacyl-CoA dehydrogenase, Δ(3,5),Δ(2,4)-dienoyl-CoA isomerase, the elongation factor 1-δ, the 26S protease regulatory subunit 7 and the transitional endoplasmic reticulum ATPase were up-regulated, while the 78 kDa glucose-regulated protein, peroxiredoxin 4 and plakoglobin were down-regulated. The expression change of Cbr1 which is closely related to diabetic complications was further validated by western blotting. Our results and those of the bioinformatics analysis suggest that oxidative stress, the Wnt pathway, fatty acid degradation and glucose transport may be closely related to T1DM.

null

pubmed

null

2,011

null

21,567,075

International journal of molecular medicine

2205.09907

R. Jagannathan

Sameen Ahmed Khan (Department of Mathematics and Sciences, College of Arts and Applied Sciences, Dhofar University, Salalah, Sultanate of Oman) and Ramaswamy Jagannathan (The Institute of Mathematical Sciences, Chennai, India)

A new matrix representation of the Maxwell equations based on the Riemann-Silberstein-Weber vector for a linear inhomogeneous medium

50 pages. The previous version has been substantially revised. Journal Reference: Results in Optics (in press)

Results in Optics 17 (2024) 100747

10.1016/j.rio.2024.100747

null

math-ph math.MP physics.class-ph physics.optics

http://creativecommons.org/licenses/by/4.0/

We derive a new eight dimensional matrix representation of the Maxwell equations for a linear homogeneous medium and extend it to the case of a linear inhomogneous medium. This derivation starts ab initio with the Maxwell equations and uses arguments based on the algebra of the Pauli matrices. This process leads automatically to the matrix representation based on the Riemann-Silberstein-Weber (RSW) vector. The new representation for the homogeneous medium is a direct sum of four Pauli matrix blocks. This aspect of the new representation should make it suitable for studying the propagation of electromagnetic waves in a linear inhomogeneous medium adopting the techniques of quantum mechanics treating the inhomogeneity as a perturbation. The new representation is used to rederive the Mukunda-Simon-Sudarshan matrix substitution rule for transition from the Helmholtz scalar wave optics to the Maxwell vector wave optics.

[ { "created": "Fri, 20 May 2022 00:28:14 GMT", "version": "v1" }, { "created": "Sun, 7 Aug 2022 06:06:58 GMT", "version": "v2" }, { "created": "Wed, 16 Oct 2024 16:25:33 GMT", "version": "v3" } ]

2024-11-06

[ [ "Khan", "Sameen Ahmed", "", "Department of Mathematics and Sciences, College of\n Arts and Applied Sciences, Dhofar University, Salalah, Sultanate of Oman" ], [ "Jagannathan", "Ramaswamy", "", "The Institute of Mathematical Sciences, Chennai,\n India" ] ]

Fri, 20 May 2022 00:28:14 GMT

2022-05-20T00:28:14Z

[ "physics", "math" ]

arxiv

null

1604.05404

Wujiang Lou

Repo Haircuts and Economic Capital: A Theory of Repo Pricing

45 pages, 5 figures, 9 data tables

null

q-fin.PR q-fin.RM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

A repurchase agreement lets investors borrow cash to buy securities. Financier only lends to securities' market value after a haircut and charges interest. Repo pricing is characterized with its puzzling dual pricing measures: repo haircut and repo spread. This article develops a repo haircut model by designing haircuts to achieve high credit criteria, and identifies economic capital for repo's default risk as the main driver of repo pricing. A simple repo spread formula is obtained that relates spread to haircuts negative linearly. An investor wishing to minimize all-in funding cost can settle at an optimal combination of haircut and repo rate. The model empirically reproduces repo haircut hikes concerning asset backed securities during the financial crisis. It explains tri-party and bilateral repo haircut differences, quantifies shortening tenor's risk reduction effect, and sets a limit on excess liquidity intermediating dealers can extract between money market funds and hedge funds.

[ { "created": "Tue, 19 Apr 2016 01:58:28 GMT", "version": "v1" }, { "created": "Sat, 29 Oct 2016 16:33:17 GMT", "version": "v2" }, { "created": "Sat, 4 Jul 2020 15:57:48 GMT", "version": "v3" } ]

2020-07-07

[ [ "Lou", "Wujiang", "" ] ]

Tue, 19 Apr 2016 01:58:28 GMT

2016-04-19T01:58:28Z

[ "q-fin" ]

arxiv

null

Natural {SQL}: Making {SQL} Easier to Infer from Natural Language Specifications

null

Addressing the mismatch between natural language descriptions and the corresponding SQL queries is a key challenge for text-to-SQL translation. To bridge this gap, we propose an SQL intermediate representation (IR) called Natural SQL (NatSQL). Specifically, NatSQL preserves the core functionalities of SQL, while it simplifies the queries as follows: (1) dispensing with operators and keywords such as GROUP BY, HAVING, FROM, JOIN ON, which are usually hard to find counterparts in the text descriptions; (2) removing the need of nested subqueries and set operators; and (3) making the schema linking easier by reducing the required number of schema items. On Spider, a challenging text-to-SQL benchmark that contains complex and nested SQL queries, we demonstrate that NatSQL outperforms other IRs, and significantly improves the performance of several previous SOTA models. Furthermore, for existing models that do not support executable SQL generation, NatSQL easily enables them to generate executable SQL queries, and achieves the new state-of-the-art execution accuracy.

null

acl

inproceedings

gan-etal-2021-natural-sql

Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau

nov

2,021

Punta Cana, Dominican Republic

Association for Computational Linguistics

https://aclanthology.org/2021.findings-emnlp.174/

Gan, Yujian and Chen, Xinyun and Xie, Jinxia and Purver, Matthew and Woodward, John R. and Drake, John and Zhang, Qiaofu

Findings of the Association for Computational Linguistics: EMNLP 2021

2030--2042

null