Author(s)

ALIREZA TORAN POUR, SARA MORIDPOUR, RICHARD TAY, ABBAS RAJABIFARD

Abstract

Every year, about 19 percent of vehicle-pedestrian crashes in Melbourne metropolitan area, Australia, involves pedestrians with less than 18 years of age or school-aged pedestrians. This paper aims to identify contributing factors on vehicle-pedestrian crash severity of this age group. Reasonable walking distance to schools is applied in Geographic Information Systems (GIS) to identify vehicle-pedestrian crashes around schools. Then Boosted Decision Tree (BDT) and Cross-Validation (CV) technique are applied to explore significant factors. Results show that the distance of pedestrians from school is a significant factor on vehicle-pedestrian crash severity for this age group. This result could assist in identifying safe distance and safe zone around schools. Furthermore, Public health indicators such as income and commuting type from or to school found as other contributing factors to this crash type.

Keywords

vehicle-pedestrian crash, school-aged, BDT, Melbourne, GIS