Using Data Mining Algorithms to Analyse Benefits of Traffic Enforcement Cameras

The United Nations Generel Assembly introduced the decade 2011 to 2020 as a decade of reducing the traffic fatalities. This organization, in different resolutions, has urged all the countries to pay their especial attention to reduce the road accidents. The speed factor plays a key role in reducing the risk of the road accident; and one of the deterrence factors for unauthorized speed of cars is the traffic enforcement camera of the traffic police NAJA. In recent years a major increase in use of such devices for reducing violocations and accidents in suburban and urban highways is observed. The aim of this study is to analyse the effectiveness of traffic enforcement cameras of the taraffic police in the Tehran-Karaj highway. This is performed using data mining techniques such as modeling with through time series and regression.
First, using descriptive analysis, we study the process of violations along with the number of active cameras in the highway. In the inferential analysis step, we analyse the violations by modeling time series and non-liner power regression. In applying assessment data, the models fitted by time series had less prediction errors then their equivalent regression models. In the prediction of year 1393, the violation rate has reduced by 32% in Tehran-Karaj highway, compared to the violations in the same period of 1392. This observation along with other analysis results shows the effectiveness of cameras in the period of study. The results of analysis and mining of traffic enforcement camera data can be used in improving urban management and planning with the goal of reducing violations and accident.