Modeling and analysis of regular banking based on improved support vector machines

Abstract

In view of the low efficiency of the bank's regular financial products, an improved support vector machine (SVM) algorithm is proposed. Combining SVM algorithm with grid search algorithm (GS) and K-CV, named GS-SVM, according to the parameter distribution law of SVM algorithm, K-CV and GS are used to select the model parameters, which improves the accuracy of the model. And GS-SVM algorithm is applied to the customer data mining of a bank's fixed-term financial products, and the validity of the SVM model based on the optimal parameters is verified in the bank's fixed-term financial customer data, which effectively reduces the customer range of bank push-subscribe time deposits and improves the efficiency of bank financial products.