Reinforcement distribution for fuzzy classifiers: a methodology to extend crisp algorithms

被引：1

作者：

Bonarini, A ^{[1
]}

机构：

[1] Politecn Milan, Dept Elect & Comp Engn, PM AI & Robot Project, I-20133 Milan, Italy

来源：

1998 IEEE INTERNATIONAL CONFERENCE ON EVOLUTIONARY COMPUTATION - PROCEEDINGS | 1998年

关键词：

Learning Classifier Systems; Reinforcement Learning; Fuzzy Classifier Systems; Q-learning; TD-lambda;

D O I：

10.1109/ICEC.1998.700130

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Fuzzy Classifier Systems (FCS) implement a mapping fr om real numbers to real numbers, through fuzzy interpretation of input and output. Reinforcement Learning (RL) algorithms can be succesfully applied to develop learning FCS analogously to what can be done with Learning Classifier Systems (LCS). We motivate this approach and we present; a methodology to extend straightforwardly reinforcement distribution algorithms originally designed for crisp input. and output to fully exploit the features of FCS.

引用

页码：699 / 704

页数：6