Study On The Prediction Of Automobile Insurance Demand Based On BP Neural Network

In this paper, based on Chinese data of 1997-2013, selected the auto insurance premium income as the explained variable, selected the cost of car insurance, car ownership, gross domestic product, the consumer price index, number of traffic accidents, road area per capita, the level of urbanization such variables as explanatory variables, used the BP neural network to fit the relationship between insurance premium income and the nine explanatory variables. In this study, after making the BP neural network model, through the training, used the trained network to predict our auto insurance premium income, at the same time, this paper also used the other traditional time series prediction method to predict our auto insurance premium income. This paper found that: used these nine variables, including the cost of car insurance, car ownership, gross domestic product, the consumer price index, the traffic accident, road area per capita, the level of urbanization and the level of education as the input layer, and used the BP neural network model can make a good prediction of the insurance premium income and the average prediction error is 2.5%, at the same time, the prediction accuracy of BP neural network model is more effective than traditional prediction method for time series prediction.