PARAMETER-ESTIMATION OF DEPENDENCE TREE MODELS USING THE EM ALGORITHM

被引：41

作者：

RONEN, O ^{[1
]}

ROHLICEK, JR ^{[1
]}

OSTENDORF, M ^{[1
]}

机构：

[1] BBN HARK SYST CORP,CAMBRIDGE,MA 02138

来源：

IEEE SIGNAL PROCESSING LETTERS | 1995年 / 2卷 / 08期

关键词：

D O I：

10.1109/97.404132

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

A dependence tree is a model for the joint probability distribution of an n-dimensional random vector, which requires a relatively small number of free parameters by making Markov-like assumptions on the tree. In this letter, we address the problem of maximum likelihood estimation of dependence tree models with missing observations, using the expectation-maximization algorithm. The solution involves computing observation probabilities with an iterative ''upward-downward'' algorithm, which is similar to an algorithm proposed for belief propagation in causal trees, a special case of Bayesian networks.

引用

页码：157 / 159

页数：3

共 6 条

[1] APPROXIMATING DISCRETE PROBABILITY DISTRIBUTIONS WITH DEPENDENCE TREES [J].