On-line learning with restricted training sets: An exactly solvable case

DOI10.1088/0305-4470/32/18/308MaRDI QIDQ4951084zbMATH OpenOpenAlexFDO

Authors H. C. Rae, Peter Sollich, A. C. C. Coolen

Publication date 1 May 2000

Published in Journal of Physics A: Mathematical and General (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/cond-mat/9811231

zbMATH Keywords

dynamics of on-line Hebbian learninglarge perceptrons

Mathematics Subject Classification ID

Computational learning theory (68Q32) Learning and adaptive systems in artificial intelligence (68T05) Neural nets applied to problems in time-dependent statistical mechanics (82C32)

Abstract: We solve the dynamics of on-line Hebbian learning in large perceptrons exactly, for the regime where the size of the training set scales linearly with the number of inputs. We consider both noiseless and noisy teachers. Our calculation cannot be extended to non-Hebbian rules, but the solution provides a convenient and welcome benchmark with which to test more general and advanced theories for solving the dynamics of learning with restricted training sets.

Recommendations

• Dynamics of online Hebbian learning with structurally unrealizable restricted training sets.
• Supervised learning with restricted training sets: a generating functional analysis
• scientific article; zbMATH DE number 1569113
• scientific article; zbMATH DE number 1329965

Cited in

(10)

• On-line learning and the metrical task system problem
• Dynamics of online Hebbian learning with structurally unrealizable restricted training sets.
• scientific article; zbMATH DE number 1329965 (Why is no real title available?)
• scientific article; zbMATH DE number 1181887 (Why is no real title available?)
• Supervised learning with restricted training sets: a generating functional analysis
• On-line learning and generalization in coupled perceptrons
• scientific article; zbMATH DE number 1569113 (Why is no real title available?)
• scientific article; zbMATH DE number 1569114 (Why is no real title available?)
• scientific article; zbMATH DE number 1569117 (Why is no real title available?)
• Nonlinear Hebbian training of the perceptron

This page was built for publication: On-line learning with restricted training sets: An exactly solvable case

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4951084)