EL MÉTODO DE BAYES APLICADO EN LA IDENTIFICACIÓN DE LAS LOCALIZACIONES MÁS PROPENSAS A SUFRIR ACCIDENTES DE TRÁNSITO

Abstract

The frequency of and the reason for accidents in a particular location are random variables, variables whose true value cannot be predicted with absolute certainty. This means that the process of identification of the locations most prone to suffer traffic accidents (APL - accident-prone locations) is subject to uncertainty. Normally the identification of the APL is made with simple statistical techniques, such as the method of confidence intervals and quality control. Nevertheless, due to the limitations of these methods, the use of more elaborated methodologies, like Bayesian analysis, is advised. This type of analysis takes into account the statistical phenomena of regression to the mean and the unrelated effects. Even though Bayesian analysis is a statistically superior method, its use in the field of engineering, such as analysis of traffic accidents, is not as widespread as the use of this methodology in the field of medicine. The main reason for this is not only the complexity of the method, but also that is requires often unavailable technical support. The objective of this paper is to present a practical example of the application of the Bayesian method in the identification of the accident-prone locations (APL) in the city of Piura .