A Personal Credit Evaluation Method Based On Hybrid Data Mining Model

In the credit industry,data mining technologies have numerously been applied to the customer personal credit evaluation.One of the most popular data mining technologies is classification method.Previous studies have pointed out the use of feature selection(FS)and ensemble classifier can effectively improve the performance of bank customer's personal credit evaluation system.In this area,the main problem that needs to be studied is how to optimize the performance of the hybrid data mining model of feature selection algorithm and ensemble classifier through the coordination parameter optimization.In the thesis,a hybrid data mining model is proposed based on three-stage experiments,which can achieve the optimal combination of feature selection algorithm and ensemble classifier.The first stage is credit data acquisition and preprocessing.In second stage,the experiment will perform four feature selection algorithms on the dataset,including genetic algorithm(GA),principal component analysis(PCA),information gain ratio,and relief attribute evaluation function.In the thesis,parameters setting of FS methods is based on the classification accuracy resulted from the implementation of the support vector machine(SVM)classification algorithm.After choosing the appropriate model for each selected feature,they are applied to the base and ensemble classification algorithms.In this stage,the best FS algorithm with its parameters setting is indicated for the modeling stage of the proposed model.In the final stage,the classification algorithms are employed on four data sets that are processed by the feature selection algorithm.The experimental results show that the PCA algorithm is superior to other feature selection algorithms in the second stage.In the third stage,the classification results show that artificial neural network(ANN)and AdaBoost(Adaptive Boosting)have higher classification accuracy.Ultimately,this thesis proposes and validates a personal credit evaluation method based on hybrid data mining model,which is operative and robust in the application of credit evaluation.