Machine Learning Based Beijing Housing Price Prediction System

With rapid socioeconomic development, the real estate market faces dual challenges of increasing homebuying pressure and rising investment risks. As a unique commodity combining residential attributes and asset value, the formation mechanism of real estate prices has grown increasingly complex. It is not only influenced by multiple factors but also exhibits significant nonlinear characteristics in its overall trend. As a first tier city in China, Beijings housing prices are influenced by various factors, including economic policies, geographical location, educational resources, and transportation accessibility. Accurately forecasting housing price changes holds significant importance for homebuyers, investors, real estate developers, and government policymakers. This project aims to leverage machine learning and data analytics to propose a multi source data fusion framework. By collecting local environmental data such as air quality and noise pollution, it breaks the boundaries of traditional real estate datasets. For the first time, dynamic environmental indicators are incorporated into the housing price evaluation system. This approach systematically quantifies the impact mechanism of human living environment quality on real estate value. A baseline model, XGBoost, is constructed to handle high dimensional features, establishing a model capable of predicting Beijing housing prices. This assists users in understanding price trends and supports decision making. Zihao Wang | Hao Zhang | Tianshuang Han | Xiuyi Yang | Yinuo Liu "Machine Learning-Based Beijing Housing Price Prediction System" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-9 | Issue-5 , October 2025, URL: https://www.ijtsrd.com/papers/ijtsrd97504.pdf Paper URL: https://www.ijtsrd.com/computer-science/artificial-intelligence/97504/machine-learningbased-beijing-housing-price-prediction-system/zihao-wang