Hierarchical mixtures of experts and the EM algorithm

DMCA

Hierarchical mixtures of experts and the EM algorithm (1993)

Cached

Download Links

by Michael I. Jordan , Robert A. Jacobs

Citations:

883 - 21 self

BibTeX

@MISC{Jordan93hierarchicalmixtures,
    author = {Michael I. Jordan and Robert A. Jacobs},
    title = {Hierarchical mixtures of experts and the EM algorithm},
    year = {1993}
}

OpenURL

Abstract

We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the mixture components are generalized linear models (GLIM’s). Learning is treated as a max-imum likelihood problem; in particular, we present an Expectation-Maximization (EM) algorithm for adjusting the parame-ters of the architecture. We also develop an on-line learning algorithm in which the pa-rameters are updated incrementally. Comparative simulation results are presented in the robot dynamics domain.

Keyphrases

hierarchical mixture em algorithm on-line learning algorithm linear model tree-structured architecture statistical model mixture component max-imum likelihood problem mixture coefficient com-parative simulation result hi-erarchical mixture model supervised learning robot dynamic

DMCA