English

Field Theoretical Analysis of On-line Learning of Probability Distributions

Disordered Systems and Neural Networks 2009-01-23 v1 adap-org High Energy Physics - Theory Adaptation and Self-Organizing Systems Data Analysis, Statistics and Probability

Abstract

On-line learning of probability distributions is analyzed from the field theoretical point of view. We can obtain an optimal on-line learning algorithm, since renormalization group enables us to control the number of degrees of freedom of a system according to the number of examples. We do not learn parameters of a model, but probability distributions themselves. Therefore, the algorithm requires no a priori knowledge of a model.

Keywords

Cite

@article{arxiv.cond-mat/9911474,
  title  = {Field Theoretical Analysis of On-line Learning of Probability Distributions},
  author = {Toshiaki Aida},
  journal= {arXiv preprint arXiv:cond-mat/9911474},
  year   = {2009}
}

Comments

4 pages, 1 figure, RevTex