Research On Early Warning Model Of Personal Credit Risk Of Commercial Bank Based On Neural Network Optimization Algorithm

Since China's reform and opening up,people's living standards have been improving,and people's ideas have gradually changed.Using credit to improve their living conditions has become an important way of life for Chinese residents.With the continuous expansion of China's personal credit market,commercial banks,as an important part of China's financial sector,have encountered great challenges in how to effectively manage personal credit risk.How to establish an efficient early warning model of personal credit risk to evaluate the personal credit risk correctly is a difficult problem that commercial banks need to face in the new situation.This paper starts from the current management situation of personal credit risk in China's commercial banks,on the basis of previous studies,by optimizing the personal credit risk early warning model,constructs a personal credit risk early warning mechanism system with the personal credit risk early warning model as the core.This paper mainly includes the following aspects: first,it introduces the related concepts of personal credit risk early warning of commercial banks and the related literature review.Second,from the perspective of the causes of personal credit risk of commercial banks,combined with the relevant characteristics and development status of personal credit market in China,this paper analyzes the current situation of personal credit risk management of commercial banks in China and the related problems in management.Thirdly,based on the research of the early warning mechanism of personal credit risk at home and abroad,combined with the characteristics of China's commercial banks in personal credit business,establish the indicator system of the early warning mechanism of personal credit risk of commercial banks.Fourth,establish the early warning model of personal credit risk of commercial banks,evaluate the personal credit risk of commercial banks,and select support vector machine as the basic model by comparing the characteristics of various personal credit risk models and combining the characteristics of personal credit business of commercial banks in China.Based on the data from a cooperative survey of a commercial bank in Jiangxi Province,a personal credit risk early warning model of commercial banks is established under the support vector machine model.Then,the optimization of parameters of the support vector machine model is compared and studied.The cross validation method,PSO particle swarm optimization algorithm and neural networkalgorithm are used to analyze the same sample,and the comparative study is carried out In the framework of support vector machine model,the neural network optimization algorithm has higher accuracy than the interactive verification optimization method and PSO Particle Swarm Optimization Algorithm in parameter optimization of support vector machine model.Fifthly,starting from the department structure and system structure of the early-warning mechanism and the details that need to be paid attention to in the actual management,this paper systematically expounds the operation relationship between the departments of the early-warning mechanism of personal credit risk of commercial banks and the operation relationship between each system,and systematically expounds the operation mechanism of the early-warning mechanism of personal credit risk of commercial banks,so as to provide the construction of personal credit for commercial banks Relevant suggestions on risk early warning mechanism.In this paper,empirical research shows that under the support vector machine model as the basic model of commercial personal credit risk early warning,the personal credit risk early warning model of commercial banks based on neural network optimization algorithm is better than the personal credit risk early warning model of commercial banks based on cross validation method in the accuracy of model prediction.Moreover,when using neural network optimization algorithm to select parameters,the speed of improving accuracy is much faster than that of cross validation and particle swarm optimization,and the speed of finding the optimal combination of parameters is greatly improved.The early warning model of personal credit risk of commercial banks studied in this paper can make a good prediction,lay a good foundation for the early warning mechanism of personal credit risk of commercial banks,and play an effective role for the risk control personnel of banks in the risk management of borrowers..