Unsupervised feature selection using a neuro-fuzzy approach

A neuro-fuzzy methodology is described which involves connectionist minimization of a fuzzy feature evaluation index with unsupervised training. The concept of a ¯exible membership function incorporating weighed distance is introduced in the evaluation index to make the modeling of clusters more appropriate. A set of optimal weighing coecients in terms of networks parameters representing individual feature importance is obtained through connectionist minimization. Besides, the investigation includes the development of another algorithm for ranking of di€erent feature subsets using the aforesaid fuzzy evaluation index without neural networks. Results demonstrating the e€ectiveness of the algorithms for various real life data are provided. Ó 1998 Published by Elsevier Science B.V. All rights reserved.

• Type: Journal
• Year: 1998
• Where: PRL
• Authors: Jayanta Basak, Rajat K. De, Sankar K. Pal
• Keywords: Connectionist Minimization, Evaluation Index, Fuzzy Feature Evaluation, PRL 1998