Prediction of vehicular accident fatalities in Bogota:

During the year 2022, more than 8,000 people died in traffic accidents in Colombia, 80% of which involved motorcycles or pedestrians. Although authorities currently have aggregate accident prediction models, knowing that a certain number of fatal accidents will occur is useless in terms of prevention if the information cannot be narrowed down to a "where". This paper proposes a simple model that uses the information of the 2021 accident location in Bogota to derive the probability of occurrence of fatal accidents in the capital city. It also uses Machine Learning models to develop a predictive model of whether a traffic accident will result in a fatal accident. This would offer the Traffic Department a tool for the efficient use of resources such as civilian traffic agents to prevent such accidents. It is found that the highest probability of fatal accidents occurs on secondary roads, especially in the southwest and northwest of Bogota. Likewise, among the characteristics most associated with fatal accidents are pedestrian accidents, the existence of gaps, lack of signaling and speeding on secondary roads.