Study of machine learning techniques for pedestrian dynamics simulation models

Pedestrian modeling has made significant progress, studying the pedestrian as an individual and also their behavior in the face of obstacles in their environment, from confined access spaces to the interaction of the movement of a large number of other people. In this paper we wish to explore hybrid models that allow simulating pedestrian dynamics supported by machine learning techniques to make. The results of previous work, together with the experiments of this work, are favorable for advancing toward a methodology that can incorporate the development of pedestrian dynamics simulation based on machine learning models.