基于XGBoost算法的城市热点区域房价预测——以南京江北新区为例  

House Price Prediction in Urban Hotspot Areas Based on XGBoost Algorithm:Taking Nanjing Jiangbei New District as an Example

作　　者：朱海煜 王志杰[1] 叶灿灿 ZHU Haiyu;WANG Zhijie;YE Cancan(College of Civil Engineering,Nanjing Forestry University,Nanjing 210037,China)

机构地区：[1]南京林业大学土木工程学院,江苏南京210037

出　　处：《建筑经济》2022年第S02期433-437,共5页Construction Economy

摘　　要：房价牵动社会金融体系稳定乃至整个宏观社会的可持续发展,进行房价预测研究对于个人消费者、房地产开发商以及国家宏观调控部门均具有重要意义。本文运用细粒度房价数据、POI数据限定热点区域并计算区域房价;以消费者预期为切入点,采集网络搜索数据,通过XGBoost算法训练模型拟合热点区域房价并作出预测。结果表明,借助网络可以获取精度更高、更加灵活的数据,以网络搜索数据量化消费者预期进行房价预测研究其准确性符合预期。The housing price affects the stability of the social financial system and even the sustainable development of the whole macro society.The research on housing price prediction is of great significance to individual consumers,real estate developers and national macro-control departments.This paper uses finegrained house price data and POI data to define hot areas and calculate regional house prices;Taking consumer expectations as the starting point,collect network search data,and use XGBoost algorithm training model to fit the house prices in hot areas and make predictions.The re sults show that we can obtain more accurate and flexible data with the help of the network,and the accuracy of the research on housing price prediction by quantifying consumer expectations with network search data is in line with expectations.

关 键 词：房价预测 XGBoost POI 网络搜索数据 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] F293.3[自动化与计算机技术—控制科学与工程]