Regresar

Modeling of a Vehicle Accident Pbkp_rediction System Based on a Correlation of Heterogeneous Sources

Abstract:

Statistics affirm that traffic accidents are the main cause of death in developing countries. The indicators are alarming, so governments, manufacturers, and researchers have been looking for solutions to mitigate them. Despite all efforts to face this problem, the number of victims remains high. A significant percentage of traffic accidents are caused by external factors, so the search for solutions that use information from multiple sources is crucial. This article presents a traffic accident pbkp_rediction system based on heterogeneous sources using data mining techniques and machine learning algorithms. The development of this system includes the following tasks: collecting information from different sources, performing cluster analyses and feature selection, generating new datasets, performing machine learning algorithms to define accident rates, and sending traffic rate levels to the vehicles. For this article, we focused on performing cluster analyses to determine high-risk clusters that identify drivers with risky driving patterns.