Real Estate Valuation in Buenos Aires, an Interactive Tool Development

Abstract

This thesis explores the application of machine learning models for real estate valuation in Buenos Aires, aiming to develop a user-centric tool to assist buyers and investors in making informed decisions. Traditional regression models, while useful, often fail to capture the complex, nonlinear relationships inherent in real estate data. Consequently, we employed advanced machine learning techniques, including XGBoost, Random Forest, and Support Vector Machines (SVM), selected for their robustness and efficiency in handling large datasets. Our input data consists of 64,358 property listings from the e-commerce platform Mercado Libre, obtained through a combination of Python scripts and the platform's own API.