Companies Credit Assessment Model Based On Self-Organizing Feature Map Neural Network

As the development of information technology, artificial intelligence model has been used in credit Assessment in recent years. Artificial intelligence models in common use often contain artificial neural networks decision-making trees etc. We can regard artificial neural network as a non-linear mapping from input space to output space, and it can "study" or find the relation between variables by adjusting weight and threshold. It can sort different things and resolve the problems of non-normal school or non-linear credit Assessment efficiently.Artificial neural network comprises a great deal of basic elements — neural uint connected one another. It can simulate human being's brain neural to deal with information and implement parallel treating and non-linear conversion. Artificial neural network deals with information by using information sample to train neural network, which make artificial neural network have the brain functions of memory, distinguishing and treating information.Artificial neural network has good ability of self-study ,self-adapt, association memory etc, and it avoids the complex mathematics derivation. In the case of lacking samples or parameters excursion, it can still guarantee stable outputs. So artificial neural network has been recognized highly and studied abroad by academe. Now it has been applied successfully in many fields, such as mode identification,optimizing computing etc. In this paper, artificial neural network is applied in the fields of credit Assessment.In this paper, author analyzes the background and actuality of credit Assessment, and data mining technology. At the same time, author researches the developing actuality and characteristic of ANN. On the base of study, put forward SOM neural network based companies credit Assessment model and give the standardization methods of treating with original data through data mining technology.