摘要
为了更精确快速地预测二手房房价,提出基于梯度提升决策树的房价预测模型。首先,采集最新沈阳二手房数据,对数据进行预处理;其次,基于处理后的数据和梯度提升决策树方法建立房价预测模型;最后,利用实验验证模型的有效性。实验结果显示,基于梯度提升决策树模型在拟合优度、均方根差、平均绝对误差都优于岭回归、决策树。在预测房价上具有一定的实用性。
In order to predict second⁃hand housing prices more accurately and quickly,a prediction algorithm based on the Gradient Boosting Decision Tree(GBDT)is proposed.Firstly,collect the latest second⁃hand housing data from Shenyang and pre⁃process the data.Secondly,a housing price prediction model is established based on the processed data and Gradient Boosting De⁃cision Tree method.Finally,prediction methods such as ridge regression,random forest,and linear regression were chosen as com⁃parison methods.GBDT model outperformed other methods such as ridge regression,random forest,and linear regression in terms of evaluation indicators.It has certain practicality in predicting housing prices.
作者
宋阳
Song Yang(School of Electronic and Information Engineering,Liaoning Institute of Science and Technology,Benxi 117004,China)
出处
《现代计算机》
2024年第17期81-84,共4页
Modern Computer
关键词
梯度提升决策树
房价预测
岭回归
随机森林
数据预处理
gradient boosting decision tree
housing price prediction
ridge regression
random forest
