Combined Model For The Regional Development Of Health Insurance In China

At present,the construction of China social medical security system is not per-fect,and the scope of protection is very limited.As an effective supplement to the social security system,health insurance is conducive to alleviating the burden of the government and the pressure of enterprises,meeting people's different levels of medical services,and improving the overall social Level of medical security.However,although the development of China's health insurance has achieved re-markable results,there are still some problems.Among them,the imbalance of regional development has seriously restricted the improvement of health insurance coverage,hin-dered the implementation of health insurance off-site services,and is not conducive to health insurance fully exerting its health protection role.According to the related literature research and experimental verification,select eight indexes,namely,gross domestic product(GDP),the level of social security,fi-nancial employment ratio,disposable income,education level,urbanization rate,elderly dependency ratio and life expectancy,Using K-means clustering method to redefine the development area of China's health insurance.After that,according to the seasonal characteristics of monthly premium income data of health insurance,various models were compared and analyzed,and it was found that a single prediction model could not fully reflect all the information of seasonal data,and no one had built a combi-nation model of seasonal time series model and seasonal SVR model.Therefore,after fully considering the obvious seasonal characteristics and trend characteristics in the monthly premium income of health insurance,this paper selects the combined model of SARIMA and seasonal SVR for prediction.Among them,SARIMA model is an im-proved model after considering the seasonal factors of ARIMA model.The seasonal SVR model is suitable for processing nonlinear data and has good data fitting and pre-diction ability.In this paper,the least square method is used as the criterion and the error square minimization is used as the objective function to determine the optimal weight coefficient of the combination model and obtain the final combination prediction model.The empirical study shows that SARIMA and seasonal SVR combined model have better prediction effect on seasonal data and improve the prediction accuracy of the model.