Pattern recognition for road traffic accident severity in Korea

An increasing number of road traffic accidents (RTA) in Korea has emerged a s being harmful both for the economy and for safety. An accurately estimate d classification model for several severity types of RTA as a function of r elated factors provides crucial information for the prevention of potential accidents. Here, three data-mining techniques (neural network, logistic re gression, decision tree) are used to select a set of influential factors an d to build up classification models for accident severity. The three approa ches are then compared in terms of classification accuracy. The finding is that accuracy does not differ significantly for each model and that the pro tective device is the most important factor in the accident severity variat ion.