基于LightGBM模型的二手房房价预测研究  

Research on prediction of second-hand housing price based on LightGBM model

作　　者：慕钢 张宏烈[1] 党佳俊 李广峰 MU Gang;ZHANG Honglie;DANG Jiajun;LI Guangfeng(School of Computer and Control Engineering,Qiqihar University,Qiqihar 161006,China;No.1 Middle School of Kedong County of Qiqihar,Qiqihar 164800,China)

机构地区：[1]齐齐哈尔大学计算机与控制工程学院,黑龙江齐齐哈尔161006 [2]齐齐哈尔克东县第一中学,黑龙江齐齐哈尔164800

出　　处：《高师理科学刊》2020年第12期27-31,共5页Journal of Science of Teachers＇College and University

基　　金：黑龙江省教育厅基本业务专项理工面上项目(135209234)。

摘　　要：房价预测是典型的机器学习问题,可以选择多元线性回归方法进行分析.基于LightGBM算法,以二手房房价数据作为研究依据,建立房价预测模型进行统计推断.以拟合优度R2作为评判依据,该模型与其它常用机器学习模型进行预测结果比较.程序运行结果表明,此房价预测模型具有更准确的预测效果.House price prediction is a typical machine learning problem,which can be analyzed by multiple linear regression method.Based on LightGBM algorithm,takes the second-hand housing price data as the research basis,establishes the housing price prediction model for statistical inference.Based on the goodness of fit R2,the model is compared with other machine learning models.The program running results show that the model proposed has more accurate prediction effect.

关 键 词：房价预测 机器学习 数据分析 LightGBM 预测对比 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]