Comparing Machine Learning Models for House Price Prediction: Linear Regression, Decision Tree, and Random Forest

College of Engineering, Texas A&M University, College Station, TX 77840, USA

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Predicting housing prices accurately is crucial for informed real estate decisions. This paper compares the performance of three machine learning models—Linear Regression, Decision Tree, and Random Forest— on the California Housing Prices dataset from scikit-learn. Through comprehensive data preprocessing and hyperparameter tuning using grid search with cross-validation, the paper evaluates model performance using Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Square Error (RMSE), and R-squared (R²) metrics. Experimental results show that Random Forest achieves the highest prediction accuracy with the smallest MAE (0.33), MSE (0.26), and RMSE (0.51) values, and the R² value (0.78) closest to 1. The Decision Tree model demonstrates moderate accuracy but shows signs of overfitting, while Linear Regression exhibits the lowest accuracy but best robustness with minimal difference between training and test performance. These findings suggest that ensemble methods like Random Forest are particularly effective for housing price prediction, though model choice should consider the specific trade-offs between accuracy and robustness. The study provides valuable insights for selecting appropriate models in real estate price prediction applications.