Real-Time Traffic Accident Severity Prediction Using Data Mining Technologies

With the urban transport development in the recent years, frequent traffic accidents and other problems need to be improved. Understanding the causes of traffic accidents and making an early alarm model for the driver will be crucial to solve traffic accident problems in some way. In this paper, we focus on the factors that can be collected in real-time and process the factors using data mining technologies. Finally, we evaluate the performance of different classifiers. The results show that our feature processing is effective in improving the classification accuracy and we can use the model to predict the severity of traffic accident furthermore prevent traffic accidents.