Neural network-based road accident forecasting in transportation and public management

Abstract

The development of Information and Communication Technology (ICT) has influenced transportation management in multiple ways. The application of artificial intelligence techniques has gained ground lately in many scientific sectors. In this research, artificial neural network models were constructed in order to predict data about the road accidents in the study area. Several parameters were taken into consideration in order to optimize the predictions and to build the optimal forecasting model such as the number of the neurons in the hidden layers and the nature of the transfer functions. A Feedforward Multilayer Perceptron (FFMLP) was utilized, as it is considered as one of the most suitable structures for time series forecasting problems according to the literature. The optimal prediction model was tested in the study area and the results have shown a very good prediction accuracy. The road accident predictions will help public management to adopt the appropriate transportation management strategies. © Springer Nature Switzerland AG 2019.