A hybrid genetic fuzzy neural network algorithm designed for classification problems involving several groups

✍ Scribed by Ralf Östermark

• Publisher: Elsevier Science
• Year: 2000
• Tongue: English
• Weight: 209 KB
• Volume: 114
• Category: Article
• ISSN: 0165-0114
• DOI: 10.1016/s0165-0114(98)00057-8

✦ Synopsis

We propose a multigroup classiÿcation algorithm based on a hybrid genetic fuzzy neural net (GFNN) framework. Recent results on evolutionary computation and fuzzy neural network methodology are combined to e ectively adapt the membership functions of the fuzziÿer and the defuzziÿer to the data set. Separate membership functions are deÿned for each dimension in the fuzziÿer and for each fuzzy output group in the defuzziÿer. The signal inherent in the fuzziÿer is aggregated by a suitable T -norm and transmitted to the defuzziÿer. The defuzziÿer aggregates the response, i.e., the predicted group membership, by a suitable conorm. If misclassiÿcations occur during training, the membership functions of both the fuzziÿer and the defuzziÿer are adapted by a systematic, robust procedure. The algorithm is successfully tested with real economic data. In total, the GFNN performs as good as the best of the competing methods in our test. The results suggest economically meaningful interpretations.