The Performance Evaluation and Choice of Combining Forecast Method

In order to evaluate the performance and choose combining forecast method, the paper uses two single forecasting methods, namely BP neural networks and support vector machines (SVM), to forecast the Shanghai Industrial Index, the Shanghai Commercial Index, the Shanghai Real Estate Index and the Shanghai Public Utilities Index. Then it uses BP neural-based combining forecast model and SVM-based combining forecast model to forecast the above indexes. Through evaluating the results of these forecasting methods, it is argued that choosing the method that has the best forecasting result as the combining forecast model can greatly enhance the forecast effectiveness.