Application of the Empirical Bayes Method with the Finite Mixture Model for Identifying Accident-Prone Spots

摘要: Hotspot identification (HSID) is an important component of the highway safety management process. A number methods have been proposed to identify hotspots. Among these methods, previous studies indicated that empirical Bayes (EB) method can outperform other for identifying hotspots, since EB combines historical crash records site and expected crashes obtained from a performance function (SPF) similar sites. However, SPFs are usually developed based on large sites, which may contain heterogeneity in traffic characteristic. As result, hotspot accuracy possibly be affected by SPFs, when present data. Thus, it necessary consider homogeneity roadway segments using methods. To address this problem, paper three different classification-based Rural data collected Texas were analyzed classified into groups Based modeling results dataset, found one performs better than standard as well HSID