English

Storage capacity of a constructive learning algorithm

Disordered Systems and Neural Networks 2009-10-31 v1

Abstract

Upper and lower bounds for the typical storage capacity of a constructive algorithm, the Tilinglike Learning Algorithm for the Parity Machine [M. Biehl and M. Opper, Phys. Rev. A {\bf 44} 6888 (1991)], are determined in the asymptotic limit of large training set sizes. The properties of a perceptron with threshold, learning a training set of patterns having a biased distribution of targets, needed as an intermediate step in the capacity calculation, are determined analytically. The lower bound for the capacity, determined with a cavity method, is proportional to the number of hidden units. The upper bound, obtained with the hypothesis of replica symmetry, is close to the one predicted by Mitchinson and Durbin [Biol. Cyber. {\bf 60} 345 (1989)].

Keywords

Cite

@article{arxiv.cond-mat/9909248,
  title  = {Storage capacity of a constructive learning algorithm},
  author = {Arnaud Buhot and Mirta B. Gordon},
  journal= {arXiv preprint arXiv:cond-mat/9909248},
  year   = {2009}
}

Comments

13 pages, 1 figure