A Comparison of Shanghai Housing Price Index Forecasting

We forecast Shanghai housing price index using the classical time series analysis method: auto-regressive integrated moving average (ARIMA) model and two nonparametric techniques: artificial neural networks (NN) and support vector machines (SVMs). By evaluating prediction errors, we find that NN method and SVM method are obviously superior to ARIMA for the long-term forecast. It shows that NN model and SVM model are better the ability of generalization. Our study also shows that the forecasting results ofNN method and SVM method are more accurate than ARIMA in the short-time forecast.