A New Fuzzy Learning Scheme for Competitive Neural Networks
Mohammed Madia, Abdelaziz Bouroumi · 2012
To improve the eciency of competitive neural network in solving clustering problems, the outline of a new scheme of training is presented in this paper. This technique is based on an unsupervised fuzzy competitive learning, that’s why we name it fuzzy competitive learning or FCL. Results provided by the proposed technique are compared with those obtained by other well known techniques such as LVQ, GLVQ, FLVQ and FCM. Mathematics Subject Classication: 68T10