Forecasting Road Traffic Accidents: Grey System Theory GM(1,1) and Grey Entropy Based Approach (GMEPA)

Road collision is one of the worst case scenarios involving massive damages and casualties. It has become a major concern for everyday road users as well as the government. Traffic accidents in terms of forecasting can be considered as a grey system considering the complexity and unknown influencing factors causing these accidents. Therefore, it can be analyzed using GM(1,1) since grey model has the criteria of handling limited amount of data to estimate the behavior of an unknown system. However, conventional method of GM(1,1) has several drawbacks that requires improvements in order to provide a more reliable references allowing responsible authorities to come out with strategies to prevent road accidents. In this study, we compare the results of propose method which is the hybrid Grey model with Minimize Entropy Principle Approach (GMEPA) and original grey model GM(1,1) based on the minimization of forecasting error. The data used are road traffic accidents in Malaysia from 2003 to 2016 and road traffic accidents in India from year 2002 to 2015. Mean average percentage error(MAPE) and Mean Squared Error(MSE) were calculated for both models to examine which method gives the best prediction accuracy. The results conclude that GMEPA can improve the measurement of forecasting accuracy for road accident data in India but vice versa in road accident data Malaysia where it was much preferable for the application of GM(1,1).