Determinación de zonas homogéneas fisicas urbanas mediante inteligencia artificial: caso de estudio Entrerríos, Antioquia

Abstract

This study evaluated the use of artificial intelligence (AI) to modernize the delimitation of Homogeneous Urban Physical Zones (ZHFU) in Entrerrios, Antioquia, addressing the limitations of traditional IGAC methods, such as subjectivity and inefficiency. By implementing a Random Forest model in RStudio—trained with data from Guatapé and Entrerrios—an accuracy of 95.22% was achieved in classifying key variables (topography, public services, and land use), though challenges were identified with minority classes due to data imbalance. The results demonstrated that AI optimizes cadastral management, reducing costs and processing time, but its success depends on standardized data and human oversight. The study proposes AI as a complementary tool, particularly for resource-limited municipalities in Colombia, aligning with the goals of the Multipurpose Cadastre and the National Development Plan.