The Research of RBF Neutral Network Based on GMDH Objective Clustering

In this paper, the GMDH objective cluster is applied to RBF neural network. It can solve shortcomings about that the RBF network cannot be objectively selected basis function center and improve the network training speed and accuracy rate. In order to detect the cotton contaminations, it use the improved RBF neutral network to do it. And the result shows that the improved RBF neutral network could locate contaminations in cotton precisely. That has significance in improving the quality of fabrics and reducing production costs.