A NONPARAMETRIC TEST FOR COMPARING ESTIMATORS IN MARKOV RANDOM-FIELDS

被引：2
|
作者：
CHEN, CC
机构：
[1] Institute of Computer Science, National Tsing Hua University, Hsinchu
来源：
PATTERN RECOGNITION LETTERS | 1990年 / 11卷 / 11期
关键词：
Auto-binomial model; coding method; generalized Ising model; Gibbs random field (GRF); Markov random field (MRF); maximum pseudo-likelihood method; texture; Wilcoxon rank-sum test;
D O I：
10.1016/0167-8655(90)90095-J
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
This paper establishes a nonparametric test to compare estimators in Markov random fields. It is particularly useful when the behaviors of estimators are unknown. The proposed methodology is used to compare two estimators, based on coding method and maximum pseudo-likelihood method, in the generalized Ising and auto-binomial texture models. Experiments show that the estimator based on maximum pseudo-likelihood is superior to the estimator based on coding method. © 1990.
引用
收藏
页码：765 / 770
页数：6
相关论文
共 50 条
12345