Guangzhou Housing Price Prediction Based On Machine Learning Regression Models

Abstract

Nowadays, the fluctuation of housing prices is a key concern of homeowners, the real estate market and the government. The researchers found that it was possible to predict house prices accurately by analyzing relevant attributes and using the most effective models. However, due to the impact of the COVID-19 pandemic, there is still a research gap on which model is more suitable for predicting housing prices after the outbreak of COVID-19. Therefore, this study collected the latest real estate data of Guangzhou in China through the web crawler, trained the random forest regression model with collected data, and obtained a model that could predict the housing price by inputting corresponding attributes. The MSE of the trained regression model is 2802, MAE is 534, and the determination coefficient (R²_score) is 0.89, while the MSE of the trained XGBoost regression model is 3108, MAE is 643, and the determination coefficient (R²_score) is 0.87.The random forest regression model has been shown to be more accurate at predicting Guangzhou house prices after the COVID-19 outbreak. Our paper has great potential in house price prediction under the pandemic situation.