An unsupervisedcompetitive neural network algorithm for clustering mixtures of Gaussian probability density functions is proposed. The algorithm based on centroid neural network with Bhattacharyya distance is evaluat...
详细信息
An unsupervisedcompetitive neural network algorithm for clustering mixtures of Gaussian probability density functions is proposed. The algorithm based on centroid neural network with Bhattacharyya distance is evaluated in the context of speech recognition and the results show that it can reduce the Gaussian mixtures by almost 60% over the k-means algorithm.
A method using Self-Organizing Feature Map Network to generate the rule base of fuzzy control from measured data is presented. algorithm of unsupervisedcompetitivelearning is used to speed learning and convergence o...
详细信息
ISBN:
(纸本)0780342534
A method using Self-Organizing Feature Map Network to generate the rule base of fuzzy control from measured data is presented. algorithm of unsupervisedcompetitivelearning is used to speed learning and convergence of training and reduce the degree of noise disturbance. We study consistency and completeness of fuzzy control rules and we also design a interactive utility system for fuzzy controller in order to collect sample data and examine the proposed method in computer. The fuzzy controller formed by auto-generated rule base has been applied to time-delay system in simulation experiments. The step response shows that the method of generating rule base is correct.
暂无评论