A model for identifying and ranking dangerous accident locations: a case study in Flanders |
| |
Authors: | Tom Brijs Filip Van den Bossche Geert Wets Dimitris Karlis |
| |
Affiliation: | Transportation Research Institute, Limburgs Universitair Centrum, Universitaire Campus, Gebouw D, B-3590 Diepenbeek, Belgium; Department of Statistics, Athens University of Economics and Business, 76, Patission Str., 10434 Athens, Greece |
| |
Abstract: | These days, road safety has become a major concern in most modern societies. In this respect, the determination of road locations that are more dangerous than others (black spots or also called sites with promise) can help in better scheduling road safety policies. The present paper proposes a multivariate model to identify and rank sites according to their total expected cost to the society. Bayesian estimation of the model via a Markov Chain Monte Carlo approach is discussed in this paper. To illustrate the proposed model, accident data from 23,184 accident locations in Flanders (Belgium) are used and a cost function proposed by the European Transport Safety Council is adopted to illustrate the model. It is shown in the paper that the model produces insightful results that can help policy makers in prioritizing road infrastructure investments. |
| |
Keywords: | Gibbs sampling Markov Chain Monte Carlo empirical Bayes road accidents multivariate Poisson distribution |
|
|