Regression models of highway traffic crashes: a review of recent research and future research needs

Abstract

Increasing road traffic over the years and large variation in traffic composition have, collectively, resulted in simultaneous increase in road accidents particularly in most of the developing countries like India. Thus, road accident has become a major concern and analyzing accident data has been an important look out to the analysts who are in search of having a unified method of predicting road crashes under such traffic. While, significant efforts have been made over the past few decades aimed at developing statistical models for crash prediction, most of them do not consider the effect of heterogeneity of traffic mix. This calls for an initiative which would address a methodological blueprint for developing a model which would predict road crashes with reasonable amount of accuracy. Accordingly, the paper concludes with a detailed insight of crash models used in past and direction for evolving a more compatible one particularly for such traffic.