%0 Journal Article
%T 基于机器学习的成都市住房租金预测研究
Research on Housing Rent Prediction in Chengdu Based on Machine Learning
%A 汪宇
%A 陈美
%J Computer Science and Application
%P 138-150
%@ 2161-881X
%D 2025
%I Hans Publishing
%R 10.12677/csa.2025.153066
%X 我国城市现代化与住房商品化推进中，租赁市场热度上升，但租金定价普遍依赖房东主观判断，导致信息不对称和资源浪费。科学定价分析与预测成为解决租赁市场健康发展问题的有效途径。基于此，本文采用机器学习算法对成都市多个区域的住房租金进行了影响因素分析和预测研究。首先，构建了包含丰富特征的数据集并进行了数据预处理和特征工程。接着，分别建立了基于随机森林、XGBoost、BP神经网络算法的三种住房租金预测模型，并采用网格寻优方法对模型参数进行了调优。最后，采用多种评价指标对模型的预测性能进行了评估。结果表明，房屋面积、所在行政区域、朝向和出租方式是影响租金的关键因素。XGBoost预测模型在租金预测中表现最佳，具有更小的误差和更强的泛化能力。
Amid the rapid urban modernization and housing commercialization in China, the rental market has witnessed increasing popularity. However, rent pricing largely relies on landlords’ subjective judgments, leading to information asymmetry and resource wastage. Scientific pricing analysis and prediction emerge as effective approaches to addressing the healthy development of the rental market. Based on this background, this paper employs machine learning algorithms to conduct an analysis of influencing factors and predictive research on housing rents in multiple districts of Chengdu. Firstly, a dataset with rich features was constructed, followed by data preprocessing and feature engineering. Subsequently, three rent prediction models based on Random Forest, XGBoost, and BP Neural Network algorithms were established, respectively, and the model parameters were optimized using a grid search method. Finally, multiple evaluation metrics were adopted to assess the predictive performance of the models. The results indicate that house area, administrative district, orientation, and rental method are key factors influencing rent. The XGBoost prediction model performs best in rent prediction, with smaller errors and stronger generalization ability.
%K 租金预测，
%K 机器学习，
%K 随机森林，
%K XGBoost，
%K BP神经网络
Rent Prediction
%K Machine Learning
%K Random Forest
%K XGBoost
%K BP Neural Network
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=109816