Optimization of on-line principal component analysis

DOI10.1088/0305-4470/32/22/306MaRDI QIDQ4950533zbMATH OpenOpenAlexWikidataFDO

Authors E. Schlösser, David Saad, Michael Biehl

Publication date 6 April 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/9812114

zbMATH Keywords

node-dependent learning-rateson-line principal component analysisunsupervised on-line learning algorithms

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Neural nets applied to problems in time-dependent statistical mechanics (82C32) Neural networks for/in biological studies, artificial life and related topics (92B20)

Abstract: Different techniques, used to optimise on-line principal component analysis, are investigated by methods of statistical mechanics. These include local and global optimisation of node-dependent learning-rates which are shown to be very efficient in speeding up the learning process. They are investigated further for gaining insight into the learning rates' time-dependence, which is then employed for devising simple practical methods to improve training performance. Simulations demonstrate the benefit gained from using the new methods.

Recommendations

• The dynamics of on-line principal component analysis
• On the optimality of the Oja's algorithm for online PCA
• scientific article; zbMATH DE number 1569111
• Online algorithm of coupled principal (minor) component analysis
• Online principal components analysis

Cited in

(8)

• Online Schatten quasi-norm minimization for robust principal component analysis
• Some methods for optimization of neural networks
• The dynamics of on-line principal component analysis
• scientific article; zbMATH DE number 1569111 (Why is no real title available?)
• On the optimality of the Oja's algorithm for online PCA
• Principal component analysis and optimal portfolio
• The dynamics of representation learning in shallow, non-linear autoencoders
• Optimal protocols for continual learning via statistical physics and control theory

This page was built for publication: Optimization of on-line principal component analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4950533)