Model Research on Forecast of Second-Hand House Price in Chengdu Based on XGboost Algorithm

Abstract

In order to better and more accurately study the housing price of second-hand houses, this paper analyzed and studied 35417 pieces of data captured by Chengdu HOME LINK network. Firstly, the captured data were cleaned and the characteristics were selected. Then, multiple linear regression, decision tree and XGboost models were used to fit the predicted housing price score curve for these ten factors, and finally, the optimal prediction model was selected through parameter adjustment. The experimental results show that the accuracy of XGboost prediction is the highest, and the prediction accuracy score reaches 0.9251. Compared with linear regression and decision tree model, XGboost algorithm has better generalization ability and robustness in data prediction, and also prevents overfitting phenomenon, laying a solid foundation for the subsequent second-hand house price prediction.