Housing Price Prediction Using Machine Learning Algorithm

Abstract

This study examines the effectiveness of various machine learning models, including K-Nearest Neighbors (KNN), Ridge Regression, Random Forest, and Extreme Gradient Boosting (XGBoost), in predicting housing prices, using Multiple Linear Regression as a traditional baseline for comparison. Utilizing a dataset of residential property sales in Ames, Iowa, the XGBoost model was found to significantly outperform the baseline, highlighting the efficacy of machine learning techniques in this domain. Furthermore, the study applied the TreeSHAP package to enhance the interpretability of the XGBoost model, effectively bridging the gap between prediction accuracy and interpretability. The results underscore the potential of machine learning methods, particularly XGBoost, in housing price prediction, suggesting a need for further research to validate these findings using different datasets and exploring other machine learning algorithms.