Property Price Prediction (2024)

This project builds a supervised model to estimate property prices in Buenos Aires. The dataset was cleaned and standardized, then enriched with public transportation accessibility features sourced from the Buenos Aires City Government. Feature engineering and preprocessing pipelines were implemented to ensure consistent inputs and reproducible training.

Several regressors were evaluated; XGBoost delivered the best performance with an R² ≈ 0.76, providing strong baseline accuracy. A lightweight Pygame interface allows stakeholders to explore predictions interactively and run quick “what-if” scenarios based on property attributes and transit access.