Estudio comparativo de clasificación de avance de proyectos en un marco de complejidad a través de redes neuronales y máquinas de vectores de soporte

Abstract

This work presents a comparative study of the classification of projects progression through two computational models: one based on Neural Networks and another based on Support Vector Machines, having as a reference aspects of complexity in projects according to the TOE (Technical, Organizational and Environmental framework) and a database of projects obtained from the National Planning Department (DNP) Investment Monitoring Project (SPI). Large-scale engineering projects tend to fail in cost overruns and delays due to the increase in complexity and the little importance attached to it. This has promoted new paradigms of study of complex projects and the inclusion of methods of soft systems to model the management and possible results of them. This study could be the starting point for the definition of a generalized support tool in DNP's large-scale project management.