كاربرد شبكه ي عصبي مصنوعي در مدل هاي انتخاب وسيله

NEURAL NETWORK APPROACH IN MODE CHOICE MODELING

Pricing policy is a method for controlling air pollution in urban central areas. The policy can be implemented on a daily basis, in such a way that the pollution level is kept close to its standard level. The ultimate goal of such a research can be the evaluation of a pricing policy for Tehran central zone areas by means of a set of mode choice models. Stated preference base mode choice models have been developed using the neural network concept; as an alternative to the existing choice model approach. Then, the models were compared with another set of models developed by the logit structure, based on the "percent of correct prediction" criterion. A rather marginal superiority of models based on the neural network approach can be reported.