keywords: Road Traffic Crashes, Artificial Neural Network, ARIMA Model
The severity of injuries in motor vehicle crashes is of considerable interest to policy makers and safety specialists. Traffic safety has been recognized as an important issue due to the large economic and social impacts of road crashes. The contribution of this study lies in the development of Artificial Neural Network (ANN) model to analyze traffic accident data and to predict the death and injury severity of traffic accidents based on a 5-year traffic accident records along Abeokuta – Sagamu interchange axis obtained from the Federal Road Safety Corps (FRSC) Ogun State sector command. The output of the data processing revealed that72 people were killed while 609 were injured over the 5 years period of this study. Analysis of traffic accident data was performed using ARIMA (Autoregressive Integrated Moving Average) data-mining software to develop and validate the ANN model. ARIMA is implemented in four stages; Identification, Estimation, Diagnostics checking and Forecasting. There was a total of 242 Road Traffic Crashes (RTC) along the interchange for the entire duration of the study.