Assessing dynamic neural networks for travel time prediction

Abstract

Ularly suitable for predicting variables like travel time, but has not been adequately investigated. This study compares the travel time prediction performance of three dynamic neural network topologies with different memory settings. The results show that the time-delay neural networks out-performed the other two topologies. This topology also performed slightly better than the multilayer perceptron neural networks. © 2011 Springer-Verlag.