Statistics

Authors and titles for November 2021

[476] arXiv:2111.00633 (cross-list from cs.LG) [pdf, other]

Title: Settling the Horizon-Dependence of Sample Complexity in Reinforcement Learning

Yuanzhi Li, Ruosong Wang, Lin F. Yang

Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Data Structures and Algorithms (cs.DS); Optimization and Control (math.OC); Machine Learning (stat.ML)

[477] arXiv:2111.00705 (cross-list from cs.LG) [pdf, other]

Title: Communication-Compressed Adaptive Gradient Method for Distributed Nonconvex Optimization

Yujia Wang, Lu Lin, Jinghui Chen

Comments: Accepted by AISTATS 2022 (29 pages, 11 figures, 2 tables)

Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Optimization and Control (math.OC); Machine Learning (stat.ML)

[478] arXiv:2111.00734 (cross-list from cs.LG) [pdf, other]

Title: Robust Deep Learning from Crowds with Belief Propagation

Hoyoung Kim, Seunghyuk Cho, Dongwoo Kim, Jungseul Ok

Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)

[479] arXiv:2111.00743 (cross-list from cs.LG) [pdf, other]

Title: Towards the Generalization of Contrastive Self-Supervised Learning

Weiran Huang, Mingyang Yi, Xuyang Zhao, Zihao Jiang

Comments: Accepted by ICLR 2023

Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)

[480] arXiv:2111.00792 (cross-list from math.PR) [pdf, html, other]

Title: Shift-invariant homogeneous classes of random fields

Enkelejd Hashorva

Comments: Published J. Mult. Analysis Applications

Subjects: Probability (math.PR); Applications (stat.AP)

[481] arXiv:2111.00814 (cross-list from cs.LG) [pdf, other]

Title: Statistical quantification of confounding bias in predictive modelling

Tamas Spisak

Comments: 20 pages, 7 figures. The manuscript is associated with the the python package `mlconfound`: this https URL See manuscript repository, including fully reproducible analysis code, here: this https URL

Journal-ref: GigaScience, Volume 11, 2022, giac082

Subjects: Machine Learning (cs.LG); Quantitative Methods (q-bio.QM); Machine Learning (stat.ML)

[482] arXiv:2111.00839 (cross-list from econ.TH) [pdf, other]

Title: Expanding Multi-Market Monopoly and Nonconcavity in the Value of Information

Stefan Behringer

Comments: 16 pages, no figures

Subjects: Theoretical Economics (econ.TH); Applications (stat.AP); Machine Learning (stat.ML)

[483] arXiv:2111.00870 (cross-list from cs.LG) [pdf, other]

Title: Statistical Consequences of Dueling Bandits

Nayan Saxena, Pan Chen, Emmy Liu

Comments: In Workshop on Reinforcement Learning for Education, 14th International Conference on Educational Data Mining , Paris, France, 2021

Subjects: Machine Learning (cs.LG); Statistics Theory (math.ST)

[484] arXiv:2111.00875 (cross-list from cs.LG) [pdf, other]

Title: A moment-matching metric for latent variable generative models

Cédric Beaulac

Journal-ref: Machine Learning (2023)

Subjects: Machine Learning (cs.LG); Performance (cs.PF); Methodology (stat.ME); Machine Learning (stat.ML)

[485] arXiv:2111.00929 (cross-list from cs.LG) [pdf, other]

Title: Bounds all around: training energy-based models with bidirectional bounds

Cong Geng, Jia Wang, Zhiyong Gao, Jes Frellsen, Søren Hauberg

Comments: This paper has been accepted by NeurIPS 2021

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[486] arXiv:2111.00980 (cross-list from cs.LG) [pdf, other]

Title: Mixture Proportion Estimation and PU Learning: A Modern Approach

Saurabh Garg, Yifan Wu, Alex Smola, Sivaraman Balakrishnan, Zachary C. Lipton

Comments: Spotlight at NeurIPS 2021

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[487] arXiv:2111.01131 (cross-list from cs.HC) [pdf, other]

Title: Hierarchical Decision Ensembles- An inferential framework for uncertain Human-AI collaboration in forensic examinations

Ganesh Krishnan, Heike Hofmann

Comments: 33 pages, 17 figures

Subjects: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Applications (stat.AP)

[488] arXiv:2111.01166 (cross-list from cs.LG) [pdf, other]

Title: Dynamics of Local Elasticity During Training of Neural Nets

Soham Dan, Anirbit Mukherjee, Avirup Das, Phanideep Gampa

Comments: 40 pages (single column), the experiments have been significantly improved than the previous version

Subjects: Machine Learning (cs.LG); Dynamical Systems (math.DS); Machine Learning (stat.ML)

[489] arXiv:2111.01222 (cross-list from cs.LG) [pdf, other]

Title: Kernel Deformed Exponential Families for Sparse Continuous Attention

Alexander Moreno, Supriya Nagesh, Zhenke Wu, Walter Dempsey, James M. Rehg

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[490] arXiv:2111.01262 (cross-list from math.OC) [pdf, other]

Title: Minimax Optimization: The Case of Convex-Submodular

Arman Adibi, Aryan Mokhtari, Hamed Hassani

Subjects: Optimization and Control (math.OC); Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG); Systems and Control (eess.SY); Machine Learning (stat.ML)

[491] arXiv:2111.01387 (cross-list from cs.LG) [pdf, other]

Title: Understanding Entropic Regularization in GANs

Daria Reshetova, Yikun Bai, Xiugang Wu, Ayfer Ozgur

Comments: 29 pages, 7 figures

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[492] arXiv:2111.01395 (cross-list from cs.LG) [pdf, other]

Title: Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds

Yujia Huang, Huan Zhang, Yuanyuan Shi, J Zico Kolter, Anima Anandkumar

Comments: NeurIPS 2021

Subjects: Machine Learning (cs.LG); Cryptography and Security (cs.CR); Machine Learning (stat.ML)

[493] arXiv:2111.01479 (cross-list from cs.AI) [pdf, other]

Title: Dealing With Misspecification In Fixed-Confidence Linear Top-m Identification

Clémence Réda (UP, INSERM), Andrea Tirinzoni (Scool, CNRS), Rémy Degenne (Scool, CNRS)

Comments: Virtual conference

Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Statistics Theory (math.ST)

[494] arXiv:2111.01480 (cross-list from cs.LG) [pdf, other]

Title: A derivation of variational message passing (VMP) for latent Dirichlet allocation (LDA)

Rebecca M.C. Taylor, Dirko Coetsee, Johan A. du Preez

Comments: 24 pages,not yet submitted anywhere

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[495] arXiv:2111.01482 (cross-list from cs.LG) [pdf, other]

Title: DAGSurv: Directed Acyclic Graph Based Survival Analysis Using Deep Neural Networks

Ansh Kumar Sharma, Rahul Kukreja, Ranjitha Prasad, Shilpa Rao

Comments: 16 pages, Accepted at ACML 2021

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[496] arXiv:2111.01533 (cross-list from math.OC) [pdf, other]

Title: A comparison of mixed-variables Bayesian optimization approaches

Jhouben Cuesta-Ramirez, Rodolphe Le Riche, Olivier Roustant, Guillaume Perrin, Cedric Durantin, Alain Gliere

Comments: Accepted for publication in Advanced Modeling and Simulation in Engineering Sciences, march 2022

Subjects: Optimization and Control (math.OC); Machine Learning (cs.LG); Applications (stat.AP); Computation (stat.CO); Machine Learning (stat.ML)

[497] arXiv:2111.01555 (cross-list from cs.LG) [pdf, other]

Title: Likelihood-Free Inference in State-Space Models with Unknown Dynamics

Alexander Aushev, Thong Tran, Henri Pesonen, Andrew Howes, Samuel Kaski

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[498] arXiv:2111.01635 (cross-list from cs.LG) [pdf, other]

Title: Characterizing and Understanding the Generalization Error of Transfer Learning with Gibbs Algorithm

Yuheng Bu, Gholamali Aminian, Laura Toni, Miguel Rodrigues, Gregory Wornell

Subjects: Machine Learning (cs.LG); Information Theory (cs.IT); Machine Learning (stat.ML)

[499] arXiv:2111.01732 (cross-list from cs.LG) [pdf, other]

Title: Spatio-Temporal Variational Gaussian Processes

Oliver Hamelijnck, William J. Wilkinson, Niki A. Loppi, Arno Solin, Theodoros Damoulas

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)

[500] arXiv:2111.01743 (cross-list from cs.LG) [pdf, other]

Title: Designing Inherently Interpretable Machine Learning Models

Agus Sudjianto, Aijun Zhang

Comments: arXiv admin note: text overlap with arXiv:2011.04041

Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)