A fuzzy based Hopfield network for partitional clustering

This paper proposes a new clustering algorithm which employs an improved stochastic competitive Hopfield network in order to organize data patterns into natural groups, or clusters, in an unsupervised manner. To overcome the problem of uncertainty for clustering, this Hopfield network employs a fuzzy based energy function. Additionally, a chaotic variable is introduced in order to escape from the local minima and gain a better clustering. By maximizing the degree of membership for each data item in a cluster using Hopfield network, we achieve a superior accuracy to that of the best existing algorithms such as optimal competitive Hopfield model, stochastic optimal competitive Hopfield network, k-means and genetic algorithm. The experimental results demonstrate the scalability and robustness of our algorithm over large datasets.