Study On Personal Credit Evaluation Based On Improved Support Vector Machine

The personal credit scoring plays an important role in the development of economic and personal consumption. However, due to the fall behind of China’s personal credit development, personal credit management tools and assessment methods can’t meet the needs of economic development, this is a serious impediment to the development of individual consumption. Therefore, the research of personal credit evaluation method has important role in promoting China’s credit system and economic.This dissertation uses the personal credit scoring as it’s research object, Introduction in detail of the background and status of the personal credit evaluation studies. Analyzed and summarized the commonly used in mathematical statistics and artificial intelligence methods of personal credit scoring. Due to the unique characteristics of personal credit assessment, these methods still exist shortcomings and deficiencies. The support vector machine learning limited only by few support vectors, avoid the dimension disaster and have good robustness. Therefore, support vector machine was applied to personal credit scoring in this paper.In this paper, by depth study of the clustering and support vector machine’s theory, the personal credit scoring model has designed and implemented, and simulation experiments to compare the model prediction accuracy and computation time, the new model is better than the traditional method. In addition, the article also gives a new credit scoring method based on support vector regression and "credit scoring" concept. First, clustering analysis of samples is given "credit scoring", and use it as output to train the support vector regression model. Finally, we use the model to predict the "credit scoring" of samples. The experiment proves that the approach not only has a good performance, but also gives quantitative predictions result, more intuitive and close to reality. Finally, given the work a summary and outlook.