An approach for on-line extraction of fuzzy rules using a self-organising fuzzy neural network

Research output: Contribution to journalArticlepeer-review

250 Citations (Scopus)
Original languageEnglish
Pages (from-to)211-243
JournalFuzzy Sets and Systems
Issue number2
Publication statusPublished (in print/issue) - 1 Mar 2005

Bibliographical note

Other Details
Fuzzy modeling of complex systems requires the extraction of a suitable collection of fuzzy rules from the available data set. This is often a difficult challenge. While neural networks have the ability to learn the fuzzy rules automatically, it is usually not possible to extract the rules from the trained neural network. The importance of this paper is that it presents a new self-organising fuzzy neural network that is able to extract fuzzy rules from the data set, and implement structure and parameter learning, with a high accuracy. The automatic addition and pruning techniques implemented ensure a compact network.

Cite this