Prediction of Casualties of Traffic Accidents in Khuzestan Province Using Dynamic Neural Network

Traffic accidents and injuries are one of the major causes of mortality and disability in developing countries. The prediction of casualties is an important factor for managers to take strategic decisions. The study investigates the prediction of casualties in Khuzestan province. This is a comparative, retrospective study that uses statistical data to estimate the number of road traffic deaths in Khuzestan province between April 2008 and March 2016, which is received monthly from the forensic medicine department of Khuzestan province, and predicts the future. In order to compare the accuracy of the prediction method, the mean percentage index of absolute error was used. In this research, NARX and NAR dynamic networking methods are used to predict the number of deaths due to accidents. According to the figures, the trend of people's deaths in crashes in Khuzestan province is decreasing and decreasing, and in 2019 it will decrease to 554 people. The absolute percentage error of the NARX method is 0.102 and 0.178 NAR, which indicates that the accuracy of the NARX method is greater than the NAR, and the predicted values are more validated using NARX. Continuing the process of reducing road deaths in Khuzestan province is indicative of improving the driving safety situation in this province. Due to the high accuracy of the NARX method, it is better to use this method to estimate accidents.