摘要
本文简要讨论了离散信息熵测度及MEE估计的基本原理,并为有效地求解离散MEE估计问题提供了新颖实用的方法。从简单普适的穷搜(ES)法出发分别建立了降维搜索(RS)算法和AI算法,从而大幅度缩小了搜索空间。
In this paper, the principles of Shannon＇s entropy measure and minimum error entropy(MEE)estimation in discrete case are briefly discussed first. Then several novel and practical methods are presented to solve discrete MEE estimation problems effectively and efficiently. Based on the simple exhaustive search(ES)method, a reduced-space search(RS)algorithm and an AI algorithm are developed respectively, which make the search space reduced dramatically.
出处
《控制与决策》
EI
CSCD
北大核心
1992年第5期342-348,共7页
Control and Decision
基金
国家自然科学基金
关键词
熵
人工智能
误差
estimation theory
entropy
artificial intelligence
search
inference
minimum error entropy estimation
