Abstract
The main objective of this study is to develop models that can forecast traffic accidents, injuries and fatalities in developing countries. Consequently, this could set the first step in enhancing traffic safety measures as well as to help planning for the future. Moreover, results of this investigation may assist decision and law makers at transportation and traffic agencies to come up with resolutions for the increasing traffic crashes and consequently loss of lives, properties and the high cost involved. The forecasting models were developed using the well-known time series analysis. Each time series model was decomposed into its components (trend, cyclical, seasonal and irregular components). In conclusion, the forecasting by time series analysis explained a significant proportion of the variability in the dependent variables (accidents, injuries and fatalities) as shown by the relatively acceptable coefficients of determination (R-2).