HYBRID NEURAL NETS CAN BE FUZZY CONTROLLERS AND FUZZY EXPERT-SYSTEMS

Authors
BUCKLEY JJ HAYASHI Y
Citation
Jj. Buckley et Y. Hayashi, HYBRID NEURAL NETS CAN BE FUZZY CONTROLLERS AND FUZZY EXPERT-SYSTEMS, Fuzzy sets and systems, 60(2), 1993, pp. 135-142
Citations number
53
Categorie Soggetti
Computer Sciences, Special Topics","System Science",Mathematics,"Statistic & Probability",Mathematics,"Computer Science Theory & Methods
Journal title
Fuzzy sets and systems → ACNP
ISSN journal
01650114
Volume
60
Issue
2
Year of publication
1993
Pages
135 - 142
Database
ISI
SICI code
0165-0114(1993)60:2<135:HNNCBF>2.0.ZU;2-A
Abstract
Given a discrete fuzzy expert system we show how to construct a hybrid neural net computationally identical to the fuzzy expert system. Give n a Sugeno, Mamdani, or expert system type of controller we build a hy brid neural net that is computationally the same as the controller. Th is improves on previous results that show a (regular) neural net can a pproximate continuous fuzzy controllers and continuous fuzzy expert sy stems to any degree of accuracy.