A methodology for predicting regional freight traffic in developing countries

The prime concern of transportation planners in developing countries is how to collect and transfer data into models as simple and as effective as possible and obtain solutions in the shortest time. In this study, a methodology which utilizes multivariate statistical analysis techniques for travel estimation is presented. The two simple and convenient techniques namely Cluster Analysis and Principal Component Analysis, are used for the evaluation of available data. The traffic demand is expressed as a function of principal components which are determined independently as a small set of variables to represent total system variability. A stepwise regression analysis is carried out to derive the traffic demand‐principal component relationship. The proposed methodology is then applied to Southeastern Anatolia Regional Development Project in Turkey.