当前位置>主页 > 期刊在线 > 信息技术 >

信息技术2020年10期

基于XGBoost 与多种机器学习方法的房价预测模型
张家棋,杜金
(重庆师范大学 数学科学学院,重庆 401331)

摘  要:房价预测问题是机器学习当中典型的回归问题,常见的算法有多元线性回归、神经网络以及基于集成学习方法的XGBoost 模型,在具体的问题中,不同的模型得到的效果也不尽相同。针对房价预测这一实际问题,对房屋的各种不同特征进行分析研究,应用了多种回归模型,并比较上述三种模型在这一问题上的表现,对不同模型的优缺点进行横向对比,对效果差异进行分析与总结。


关键词:房价预测;多元回归;神经网络;XGBoost



中图分类号:TP181;F299.23         文献标识码:A         文章编号:2096-4706(2020)10-0015-04


House Price Prediction Model Based on XGBoost and Multiple Machine Learning Methods

ZHANG Jiaqi,DU Jin

(School of Mathematical Sciences,Chongqing Normal University,Chongqing 401131,China)

Abstract:The house price prediction problem is a typical regression problem in machine learning. Common algorithms includemultiple linear regression,neural networks,and XGBoost models based on integrated learning methods. Among the specific problems,different models have different effects. Aiming at the actual problem of housing price prediction,we analyze and study various differentcharacteristics of houses,apply multiple regression models,compare the performance of the above three models on this issue,andcompare the advantages and disadvantages of different models horizontally Analyze and summarize the difference in effect.

Keywords:housing price prediction;multiple linear regression;neural networks;XGBoost


参考文献:

[1] 何晓群,刘文卿. 应用回归分析:第3 版 [M]. 北京:中国人民大学出版社,2011.

[2] GARDNER W A. Learning characteristics of stochasticgradient-descent algorithms:A general study,analysis,and critique [J].Signal Processing,1984,6(2):113-133.

[3] CHEN T Q,GUESTRIN C. XGBoost:A Scalable TreeBoosting System [J/OL].arXiv:1603.02754 [cs.LG].(2016-03-09).https://arxiv.org/abs/1603.02754.

[4] FLOREZ-LOPEZ R,RAMON-JERONIMO J M. Enhancingaccuracy and interpretability of ensemble strategies in credit risk assessment.A correlated-adjusted decision forest proposal [J].Expert Systems withApplications,2015,42(13):5737-5753.

[5] 李拓,许成乾,曹琳菲,等. 基于BP 神经网络的销售量预测 [J]. 中国新通信,2020,22(1):137.


作者简介:

张家棋(2000.07—),男,汉族,四川达州人,本科,研究方向:统计学;

杜金(1999.12—),男,汉族,四川南充人,本科,研究方向:统计学。