Credit risk evaluation for listed enterprise group based on PNN

Credit risk evaluation of listed enterprise group scientifically contributes to improving the risk management ability of commercial bank. The amount of connected transaction to total assets ratio was used as index which reflects the characteristic of group, and probabilistic neural network (PNN) was used to evaluate the credit risk of listed enterprises in Shanghai and Shenzhen A shares. The empirical results show that the new model can evaluate the credit risk of group companies effectively and has higher prediction accuracy than BPNN.