Data Mining Technology's Application In Railway's Passenger Volume Forecasting

As the railway construction expanding and investment, the economic grows and people's life quality raises fast, the railway's transportation volume raising day by day. How to scientifically forecast traffic volume is of great significance for the railway department planning and controlling. The paper's purpose is how to use data mining tools to forecast traffic volume effectively.The research way in this paper is introducing principle firstly and doing demonstration analysis secondly. At first we analyze the railway transportation situation and point out the significance and necessity of forecasting railway's passenger volume. Then we assort all the popular method of passenger capacity forecasting according to the nature,means and point of view. After that the paper explores the related theory of data mining, and made a demonstration analysis using four typical DM methods.We make a carding of passenger capacity forecasting method and choose four typical data mining model:(multi-variable linear regression model),time series model (ARIMA),neural network model(GRNN) and the least square support vector machine (SVR model) to do improving demonstration analysis with comparison with former scholar. We give the reason why choose these four models and each advantages and applied conditions. Those four models all make good result expect some lacuna. And we point out the way to do better in the future.