Financial Risk Analytics Based On GA-RBF Function Neural Network

At present,the development of non life insurance companies in China i s polarized.Most of the non life insurance companies are facing the devel opment dilemma of single business source,lack of premium growth point,and six risks of investment,underwriting,eost,retirement,liquidity and mis leading sales.At the same time,China's economy has entered a new norma 1 situation,and the slowdown in economic growth has a certain impact on the non life insurance industry,the demand for non life insurance and the decline of business growth,which all make the non life insurance compani es face huge financial risks.There fore,this paper studies the financial risk early-warning model of non life insurance companies.The thesis is divided into five chapters.The first chapter mainly introdu ces the background and significance of the non life insurance company's f inancial early warning,and puts forward the innovation point of the paper.The second chapter mainly establishes the early warning mechanism for t he financial risk of the enterprise and the non life insurance company,and puts forward the selection method of the index.The third chapter introduc es the entropy value empowerment.Method,radial basis function neural n etwork nodel and genetic algorithm.In the study of the financial risk earl y warning of non life insurance companies,this paper tries to train and pre diet the selected samples on the basis of using the entropy value method t 0 establish the weight,combining with the RBF neural network model.Ent ropy method overcomes the problem of imprecise empowerment caused b y the too strong subjectivity in the analytic hierarchy process and the fuzz y evaluation method,and can objectively reflect the weight of the index an d its evaluation value.The fourth chapter is an empirical study.Taking non life insurance companies as samples in 2013?2015,the samples are traine d with RBF neural network early warning model and GA-RBF early warn ing model after sample data processing.The results are compared with the results obtained.The results show that the prediction rate of RBF neural n etwork model based on genetic algorithm is improved by 9%;finally,the p rediction rate of the model is improved.This paper makes a summary and outlook,and puts forward some suggestions to prevent financial risks of n on life financial companies.