Robust expectation maximization learning algorithm for mixture of experts

Romina Torres, Rodrigo Salas, Hector Allende, Claudio Moraga

Resultado de la investigación: Contribución a una revistaArtículo

5 Citas (Scopus)


The Mixture of Experts (ME) model is a type of modular artificial neural network (MANN) specially suitable when the search space is stratified and whose architecture is composed by different kinds of networks which compete to learn several aspects of a complex problem. Training a ME architecture can be treated as a maximum likelihood estimation problem, where the Expectation Maximization (EM) algorithm decouples the estimation process in a manner that fits well with the modular structure of the ME architecture. However, the learning process relies on the data and so is the performance. When the data is exposed to outliers, the model is affected by being sensible to these deviations obtaining a poor performance as it is shown in this work. This paper proposes a Robust Expectation Maximization algorithm for learning a ME model (REM-ME) based on M-estimators. We show empirically that the REM-ME for these architectures prevents performance deterioration due to outliers and yields significantly faster convergence than other approaches.

Idioma originalInglés
Páginas (desde-hasta)238-245
Número de páginas8
PublicaciónLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EstadoPublicada - 2003

Áreas temáticas de ASJC Scopus

  • Informática (todo)
  • Bioquímica, genética y biología molecular (todo)
  • Ciencia computacional teórica

Huella Profundice en los temas de investigación de 'Robust expectation maximization learning algorithm for mixture of experts'. En conjunto forman una huella única.

  • Citar esto