Uncertain maximum likelihood estimation with application to uncertain regression analysis

被引:62
|
作者
Lio, Waichon [1 ]
Liu, Baoding [1 ]
机构
[1] Tsinghua Univ, Dept Math Sci, Beijing 100084, Peoples R China
基金
中国国家自然科学基金;
关键词
Uncertainty theory; Regression analysis; Maximum likelihood estimation; Imprecise observation;
D O I
10.1007/s00500-020-04951-3
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Regression analysis is a mathematical tool to estimate the relationship between explanatory variables and response variable. This paper defines a likelihood function in the sense of uncertain measure to represent the likelihood of unknown parameters. Furthermore, the method of maximum likelihood estimation is used for the parameter estimation of uncertain regression models, and the uncertainty distribution of the disturbance term is simultaneously calculated. Finally, some numerical examples are documented to illustrate the proposed method.
引用
收藏
页码:9351 / 9360
页数:10
相关论文
共 50 条