Citation

Castro Y, Kim YJ. Int. J. Crashworthiness 2016; 21(2): 104-111.

Copyright

(Copyright © 2016, Informa - Taylor and Francis Group)

DOI

10.1080/13588265.2015.1122278

PMID

unavailable

Abstract

In this study, we use three data mining classification models to detect factors with the greatest influence on car accidents. Understanding the circumstances in which the drivers and passengers are more likely to be killed or severely injured in an automobile crash is of particular concern in traffic safety. Our experimental objective is exploring the role of different factors on injury risk using a Bayesian network, decision trees and artificial neural networks. To identify relevant patterns and detect the most frequent factors involved in an accident, we conducted an experiment with road accident data, from 2010 to 2012, provided by the Driver and Vehicle Standards Agency (DVSA) of the United Kingdom. Here, we evaluate and discuss our results, which show that the three most frequent factors are light conditions, vehicle manoeuvre and road type. The investigation also found that the age of the vehicle and weather conditions had no significant influence on the degree of injury.

Language: en