Prediction of freight quantity for the multiple regression method

The multiple regression method is commonly used in the freight quantity prediction because of its simple and easy. How to establish the prediction model is a question worthy of being discussed, when we want to improve the precision according to different types of cities and independent variables. This paper analysed the character of industrial structure and discovered the rule of freight quantity and unit GDP freight quantity by the example of the city Yangzhou. The paper established two prediction models by the ways of multianalysis. The first model was the relationships between freight quantity and some independent variables including Social consumable retail turnover, GDP, the first industrial value of outputs the second industrial value of output and the third industrial value of output; The second model was the relationships between unit GDP freight quantity and other independent variables including the scale of the first industry, the scale of the second industry and the scale of the third industry. Finally, the first model was proved to be more practicality and better on application in contrast with the second one in forecasting of freight quantity.