首页>
外国专利>
BP_ADABOOST MODEL-BASED METHOD AND SYSTEM FOR PREDICTING CREDIT CARD USER DEFAULT
BP_ADABOOST MODEL-BASED METHOD AND SYSTEM FOR PREDICTING CREDIT CARD USER DEFAULT
展开▼
机译:基于BP_ADABOOST模型的信用卡用户违约率预测方法及系统
展开▼
页面导航
摘要
著录项
相似文献
摘要
A BP_Adaboost model-based method and system for predicting credit card user default, comprising: acquiring attribute data of a credit card user and normalizing to produce a training sample set and a test sample set; initializing distribution weight values of training samples, determining the structure of a BP neural network, and initializing a parameter of the BP neural network; utilizing the training samples in training a number T of BP neural network weak classifiers; acquiring a strong classifier on the basis of the number T of weak classifiers, that is, a BP-Adaboost model for use in credit card user default prediction, thus predicting on the basis of the attribute data of the credit card user on whether same will default. The BP_Adaboost model-based method and system for predicting credit card user default perform data analysis and training on the basis of credit history of bank credit card users and establish a BP_Adaboost model, thus increasing the accuracy of credit card user default prediction.
展开▼