Decision-making Model for Preventing Driving Accidents Using Data analysis- A Case Study in Kerman Province

Statistics show that Iran has a larger share in terms of driving violations andaccidents in the world. According to statistics from traffic experts, about 3percent of our nation's gross domestic product annually is spent on the effectsand consequences of traffic violations. Considering the importance of the trafficdiscussion, in this study, taking into account a number of parameters related todriving violations, including offender information, vehicle information andvehicle specifications, and modeling through tools such as decision tree andfeature selection have been addressed to investigate and recognize the offendingdrivers. The data used in this study was collected during the three months of theyear 2013 based on traffic violations in Kerman. Research innovation is basedon the study of the identification of the behavior of incidental drivers and themethodology presented in this field, which first extracted important features forcategorizing drivers violated based on the purpose, and then the clusters ofdrivers will take shape. This method is implemented on real data. The results ofthis research have the potential to be implemented as a software package forregistration and control systems for traffic violations