Machine Learning-Based Classification of House Prices: A Comparative Study

This research investigates the problem of house price classification using various machine learning methods, including Random Forest, Gradient Boosting Machines (GBM), XGBoost, Logistic Regression, and K-Nearest Neighbors (KNN). The best performing method for house price classification is Random Forest, achieving an R2 value of 85.32% and an AUC of 0.93 according to the ROC evaluation. The results demonstrate Random Forest’s effectiveness in handling non-linear and complex datasets, while o ther methods such as GBM, XGBoost, Logistic Regression, and KNN also show competitive performance. This study guides house price classification modeling and supports accurate decision-making in the house market context.