The Application And Research Of Data Mining In Risk Evaluation Of Financing Guarantee

In recent years, government efforts to support the financing guarantee industry organizations continue to increase, and financing guarantee industry gradually become one of the important strategic industry in C hina. To promote its development orderly and rapidly can promote the further development of social economy in our country. At the same time, to strengthen the analysis and prediction and systematic research of financing guarantee project risk, on the one hand, it is helpful to guide the risk management of the financing guarantee industry, and promote the optimized allocation of resources, on the other hand, it also helpful to provide scientific decision basis for financing guarantee industry policy adjustment.In this paper, the improved spectral clustering algorithm is used to construct the financing guarantee project risk assessment model, and it has better solved the traditional evaluation methods of human subjective preference or subjective judgment in the evaluation process, which make the results of the assessment more objective and impartial. And using the project risk assessment model can greatly reduce the workload and improve the efficiency of evaluation. The improved spectral clustering algorithm is using the optimization algorithm based on cosine function and Particle Swarm Optimization technology, it has solved some problems of the traditional spectral clustering algorithms, such as sensitive to initialize the scale parameter, eigenvector computation complexity is higher, and so on. Eventually it has introduced to the risk assessment of financing guarantee project, so as to make total guarantee risk be the least in the project. And compared with other algorithms, the clustering efficiency and accuracy have been improved obviously. At the same time, through the review of relevant literature and reference standards, and combined with the actual situation of Guizhou Province, the paper has designed a set of financing guarantee project risk evaluation index system using the improved gray correlation analysis method, which is under the guidance of experts. In this paper, the algorithm of financing guarantee project risk assessment model is researched in detail, and on the basis of MATLAB software tool, financing guarantee project risk assessment model for experiment is established, and through training and simulation of the model, it has verified the effectiveness and superiority of the model. Finally, the evaluation model which is researched will be applied to assess the practical application, and then a set of simple financing guarantee project risk assessment system will be designed based on this. Through the empirical analysis of the financing guarantee project of recent years, and ultimately introducing the risk assessment module to the financing guarantee financial integration service platform, which providing scientific decision support for financing guarantee agencies.