Genetic-based clustering neural networks and applications
Chengyi Sun, Hongxing Chao, Yan Sun · 2002
Maximum-likelihood clustering neural networks (MLCNNs) have some prominent advantages over many other clustering algorithms. However, there is an obvious problem in MLCNNs, namely that the initial cluster centers have a great influence on the clustering results. In this paper, genetic algorithms are combined with MLCNNs to solve the problem of the selection of initial cluster centers so that the MLCNNs can give optimal clustering results. The genetic-based MLCNNs are applied to the segmentation and understanding of images through connected components and to the analysis of stock market data. In these applications, the genetic-based MLCNNs play important roles and lead to excellent results.