Machine Learning

Authors and titles for October 2021

[576] arXiv:2110.15590 (cross-list from math.ST) [pdf, other]

Title: Adaptive Importance Sampling meets Mirror Descent: a Bias-variance tradeoff

Anna Korba, François Portier

Comments: 35 pages, 5 figures

Subjects: Statistics Theory (math.ST); Machine Learning (stat.ML)

[577] arXiv:2110.15595 (cross-list from stat.ME) [pdf, other]

Title: Cause-effect inference through spectral independence in linear dynamical systems: theoretical foundations

Michel Besserve, Naji Shajarisales, Dominik Janzing, Bernhard Schölkopf

Subjects: Methodology (stat.ME); Machine Learning (stat.ML)

[578] arXiv:2110.15739 (cross-list from cs.LG) [pdf, other]

Title: Scalable Inference in SDEs by Direct Matching of the Fokker-Planck-Kolmogorov Equation

Arno Solin, Ella Tamir, Prakhar Verma

Comments: To appear in Advances in Neural Information Processing Systems (NeurIPS 2021)

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

[579] arXiv:2110.15796 (cross-list from cs.LG) [pdf, other]

Title: Properties from Mechanisms: An Equivariance Perspective on Identifiable Representation Learning

Kartik Ahuja, Jason Hartford, Yoshua Bengio

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

[580] arXiv:2110.15900 (cross-list from cs.LG) [pdf, other]

Title: Hyperparameter Tuning is All You Need for LISTA

Xiaohan Chen, Jialin Liu, Zhangyang Wang, Wotao Yin

Comments: Accepted at NeurIPS 2021

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

[581] arXiv:2110.15954 (cross-list from cs.LG) [pdf, other]

Title: Limiting fluctuation and trajectorial stability of multilayer neural networks with mean field training

Huy Tuan Pham, Phan-Minh Nguyen

Comments: 57 pages, 5 figures, NeurIPS 2021

Subjects: Machine Learning (cs.LG); Disordered Systems and Neural Networks (cond-mat.dis-nn); Machine Learning (stat.ML)

[582] arXiv:2110.15960 (cross-list from math.ST) [pdf, other]

Title: Support Recovery with Stochastic Gates: Theory and Application for Linear Models

Soham Jana, Henry Li, Yutaro Yamada, Ofir Lindenbaum

Comments: 15 pages, 7 figures, extended the theoretical studies and numerical analyses to other probabilistic models (subgaussian, moment constraints etc.)

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