Modular general fuzzy hyperline segment neural network

Abstract

This paper describes modular general fuzzy hyperline segment neural network (MGFHLSNN) with its learning algorithm, which is an extension of general fuzzy hyperline segment neural network (GFHLSNN) proposed by Patil, Kulkarni and Sontakke (2002) that combines supervised and unsupervised learning in a single algorithm so that it can be used for pure classification, pure clustering and hybrid classification/clustering. MGFHLSNN offers higher degree of parallelism since each module is exposed to the patterns of only one class and trained without overlap test and removal, unlike in fuzzy hyperline segment neural network (FHLSNN) by U.V. Kulkami et al. (2001) leading to reduction in training time. In proposed algorithm each module captures peculiarity of only one particular class and found superior in terms of generalization and training time with equivalent testing time. Thus, it can be used for voluminous realistic database, where new patterns can be added on fly.