Identifying vehicle driver injury severity factors at highway-railway grade crossings using data mining algorithms

The purpose of this paper is to identify vehicle driver injury severity factors of highway-railway grade crossing (HRGC) accidents in order to detect interactions as well as dissimilarities among accident factors. At this aim, data mining techniques were used to analyze the interaction of multiple factors in large databases. This paper applies Classification-Regression Tree (CART) and Association Rules algorithms on the U.S. Federal Railroad Administration (FRA) HRGC accident database for the period of 2006-2013 to identify vehicle driver injury severity factors at HRGCs. Both the classification trees and the rules discovery were effective in providing meaningful insights about accident factors and their interaction. The results of the two algorithms were never contradictory. Furthermore, most of the findings of this study were consistent with the results of previous studies which used different analytical techniques, such as probabilistic models of accident injury severity. The results show that train speed, type of road vehicle, driver age and gender, position of road vehicle before accident, type of accident and highway pavement type are the key factors influencing the driver injury severity.