基于网格寻优SVR房价预测模型——以郑州市为例

SVR Housing Forecast Model Based on Grid Optimization——Zhengzhou City as an Example

分析了影响房价的因素,利用房价滞后性的特点,将某年的经济指标与下一年的房价对应.利用回归支持向量机对房价进行学习、预测,同时采用网格寻优算法计算最优的学习参数.最后以郑州市的房地产数据进行仿真实验,取得了较为理想的效果.

We analyzed the factors that affect housing prices, advantage of the characteristics of hysteresis of housing prices, then, maked learning samples corresponding between the economic factors of one year and the housing prices of the next year. The regression support vector machine （SVR） was used for houing prices learning and prediction. Furthermore grid optimization algorithm was adopted for calculating optimal parameters of learning. Finally, we carried out a simulative predution with relevant data of Zhengzhou real estate and achieved good results.