Identification of singleton fuzzy models via fuzzy hyperrectangular composite NN

MaRDI QIDQ4364607zbMATH OpenFDO

Authors Mu-Chun Su

Publication date 1 July 1998

zbMATH Keywords

fuzzy hyperrectanglesfuzzy modelsfuzzy neural networksfuzzy relationslearning

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) System identification (93B30) Fuzzy control/observation systems (93C42)

Review text

This study concentrates on a class of fuzzy models. The condition parts of the rules describing the fuzzy model are represented as fuzzy hyperboxes (hyperrectangles) \(HR_i\) defined in the input space. The rules assume the form \hskip17mm -- if \({\mathbf x}\) is \(HR_i\) then \(y\) is \(c_i\) or, more generally, they may involve a union of several hyperboxes \hskip17mm -- if \({\mathbf x}\) is \((HR_i\cup HR_j)\) then \(y\) is \(c_i\) where \(c_i\) is a numeric conclusion of the rule. A learning algorithm along with some simulation results are provided. For the entire collection see [Zbl 0874.00030].

Recommendations

• Structure identification of fuzzy model
• Fuzzy system modeling using linear distance rules
• Learning rules for a fuzzy inference model
• A simple method for identification of singleton fuzzy models

Cited in

(1)

• A simple method for identification of singleton fuzzy models

This page was built for publication: Identification of singleton fuzzy models via fuzzy hyperrectangular composite NN

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4364607)