Towards effective traffic congestion prediction in Egypt

Traffic congestion in Egypt is an issue where the aspects of the problem seem different from those of western countries. The problem is exaggerated in cities away from the capital where lack of governmental interest in such areas adds another dimension to the problem. The main objective of this paper is to investigate whether it is useful to apply known traffic congestion prediction approaches on an Egyptian highway and to analyse their performance. In order to do so, the following was performed : 1) Create an Egyptian dataset for a short term highways. 2) Apply known traffic congestion techniques such as artificial neural networks and neuro-fuzzy approaches on that dataset and to test their performance. Empirical work showed that traffic congestion could be effectively predicted using the applied methods. It also showed that error obtained using these methods outperforms related work on other datasets. Additionally, tests for assuring the goodness of fit are performed.