A metric entropy bound is not sufficient for learnability

DOI10.1109/18.335898MaRDI QIDQ4838641zbMATH OpenOpenAlexFDO

Authors Sanjeev R. Kulkarni, Ofer Zeitouni, Richard M. Dudley, T. J. Richardson

Publication date 12 July 1995

Published in IEEE Transactions on Information Theory (Search for Journal in Brave)

Full work available at URL https://semanticscholar.org/paper/285fa8b9f8243fecae27f57ae27e9779941df3c2

zbMATH Keywords

probably approximately correct learning

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Recommendations

• Entropy bounds on Bayesian learning
• Metric entropy limits on recurrent neural network learning of linear dynamical systems
• Learning entropy: multiscale measure for incremental learning
• Generalization bounds for metric and similarity learning
• The entropy in learning theory. Error estimates
• Nearest-neighbor entropy estimators with weak metrics
• Learning theory of minimum error entropy under weak moment conditions
• Entropy samplers and strong generic lower bounds for space bounded learning
• Learning theory approach to minimum error entropy criterion

Cited in

(8)

• Realizable learning is all you need
• scientific article; zbMATH DE number 67635 (Why is no real title available?)
• Scale-sensitive dimensions and skeleton estimates for classification
• Characterizing rational versus exponential learning curves
• Some contributions to fixed-distribution learning theory
• Closure properties of uniform convergence of empirical means and PAC learnability under a family of probability measures.
• Entropy samplers and strong generic lower bounds for space bounded learning
• A recent result about random metric spaces explains why all of us have similar learning potential

This page was built for publication: A metric entropy bound is not sufficient for learnability

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4838641)