This study proposes a framework of model-based hotspot identification method by applying full Bayes (FB) technique. In comparison with the state-of-the-art approach, i.e., empirical Bayes method (EB), the advantage of FB method is the capability to seamlessly integrate prior information and allavailable data into posterior distributions on which various ranking criteria could be based. Using intersection crash data collected in Singapore,an empirical analysis was conducted to evaluate the following six approaches for hotspot identification: 1) naive ranking using raw crash data; 2) standard EB ranking; 3) FB ranking using Poisson-Gamma model; 4) FB ranking using Poisson-Lognormal model; 5) FB ranking using hierarchical Poisson model; and 6) FB ranking using hierarchical Poisson (AR-1) model. The results show that 1) using the expected crash rate related decision parameters, all model-based approaches perform significantly better in safety ranking than the naive ranking method; 2) FB approach using hierarchical models significantly outperforms standard EB approach in correctly identifying hazardous sites. Although the results presented in this paper demonstrated the potential of FB approach with better safety prediction performance, thestudy is subjected to several limitations due to the use of empirical data. To further explore the application of FB approach in hotspot identification, future research expanding from this study is recommended.
Abstract