Antoniou C., Koutsopoulos H., Yannis G., “Traffic state prediction using Markov chain models”, Proceedings of the European Control Conference, EUCA, Kos, July 2007, pp. 2428-2435.

Motorway traffic management and control relies on models that estimate and predict traffic conditions. In this paper, a methodology for the identification and short-term prediction of the traffic state is presented. The methodology combines model-based clustering, variable-length Markov chains and nearest neighbor classification. An application of the methodology for short-term speed prediction in a freeway network in Irvine, CA, shows encouraging results.