Aplicación de algoritmos de machine learning para predecir el precio de una vivienda

Abstract

Abstract: This thesis project develops a housing price prediction system in Ecuador based on machine learning algorithms. Through an extensive analysis of real estate data, predictive models such as Elastic Net and XGB (Extreme Gradient Boosting) were implemented using advanced pre-processing and feature selection techniques. The CRISP-DM methodology was employed to ensure a systematic and reproducible process in data handling, model training, and evaluation. Finally, the models were integrated into a web application using the AngularJs and FlaskAPI frameworks to allow users to perform real-time queries regarding a property's estimated value based on its characteristics.