Application of RBF Neural Network to Enterprise Credit Comprehensive Evaluation

Radial basis function (RBF) neural network possesses the characteristics of fast training speed, reaching the global minimum and escaping from local minimum, et al. Under fewer samples, the RBF neural network model (RBFNN-based model) for credit comprehensive evaluation is established according to actual enterprise´s credit state. For the 6 non-training enterprises, the comparison of the calculated results of several models shows that the accuracy of the RBFNN-based model established in this paper is the highest. It is effective and suitable to apply RBFNN to enterprise credit comprehensive evaluation