Analysis of Ankara urban bus accident and determination of region risks for a multi criteria decision model

The aim of this study is to predetermine the factors that cause traffic accidents and their risk assesments, and preventing death,temporary and permanent disabilities,manpower and material damages or losses.In accordance with this purpose, for determining the reasons of the traffic accidents and establishing risk situations a decision methodology and a model in the light of these is targetted.This model contains factors of the traffic accident reasons and subfactors of these. A Fuzzy Analitic Hierarchy Process was used due to determine the regional accident risk levels and establish weights of factors and sub factors , rules were created by using CHAID which is one of the decision tree algorithm of data mining technics on the basis of determined risk values and real accident datas. Model is based on the public transport vehicles of Ankara EGO Head Office and based on it.Same model was performed for the 5 areas of the EGO corporation. Developed Model has a easy structure which will be used for other corporations or facilities with minor revisions.