[en] We present a variational Expectation-Maximization algorithm to learn probabilistic mixture models. The algorithm is similar to Kohonen's Self-Organizing Map algorithm and not limited to Gaussian mixtures. We maximize the variational free-energy that sums data log-likelihood and Kullback-Leibler divergence between a normalized neighborhood function and the posterior distribution on the components, given data. We illustrate the algorithm with an application on word clustering.
Disciplines :
Computer science
Identifiers :
UNILU:UL-ARTICLE-2011-737
Author, co-author :
Verbeek, J. J.
VLASSIS, Nikos ; University of Luxembourg > Luxembourg Centre for Systems Biomedicine (LCSB)
Kröse, B. J. A.
Language :
English
Title :
Self-Organization by Optimizing Free-Energy
Publication date :
2003
Event name :
European Symposium on Artificial Neural Networks
Event date :
2003
Main work title :
Proc. of European Symposium on Artificial Neural Networks