Housing Price Forecasting Based on Least Squares Support Vector Regression with Particle Swarm Optimization

Least squares support vector regression with particle swarm optimization algorithm for housing price forecasting in the paper,and particle swarm optimization algorithm is applied to determine the parameters of relevance vector machine.Finally,housing price data of Beijing city from 2011-6-1 to 2011-6-30 are employed to study the performance of the proposed PSO-LSSVR method. The proposed PSO-LSSVR method with the 3~8 input nodes respectively are trained to find the optimal number of input nodes of the prediction models in the paper. In order to testify the better prediction performance compared with traditional support vector regression algorithm,the traditional support vector regression algorithm is applied to predict housing price of Beijing city.It can be seen that the testing results of the PSO-LSSVR model and traditional SVR model by using 5 input nodes have the best prediction effects and the prediction results for housing price of the PSO-LSSVR model are more excellent than those of the traditional SVR model.