Future Proofing Real Estate: Machine Learning for Price Predictions

Abstract

Predicting housing prices accurately is a real-world problem with huge implications in real estate planning, investment decisions, and policy-making. This research focuses on building a regression model using the XGBoost algorithm to estimate house prices based on the California Housing dataset. The dataset includes key features such as median income, average number of rooms, and population density, which all influence housing prices. Through data preprocessing, visualization, and training, we explored the relationship between these features and house prices. The model was evaluated using metrics like R-squared and Mean Absolute Error, showing strong performance and reliable predictions. Overall, this study highlights the effectiveness of gradient boosting models in real estate price forecasting and opens doors for future applications in automated valuation systems and smart housing analytics.