The Risk Assessment of Bank Based on Multiple Correlation Analysis and Three-Valued Logic Neuron Model

Most of questions have the uncertainty and the illegibility. Taking account of the insufficiency of two value logic neuron, the paper puts forward three-valued logic neural network, based on the multiple correlation analysis, which have been successfully used in the bank risk assessment. The paper, firstly has constructed the multiple correlation model. Secondly, puts forward the three-valued logic neural network and explains its basic operations such as inference rules. Finally, combining of the bank risk assessment and using the multiple correlation analysis and has determined the evaluation index and constructed an assessment of the neural network model. The result indicates that this model can be used in three-valued connection rules to realize three-valued logic reasoning. This has provided one powerful method for the uncertainty knowledge´s expression and the inference.