Traffic speed prediction in the Lyon area using DCRNN
Résumé
In this paper, we deal with the problem of traffic speed prediction, i.e., forecast- ing, based on relevant historical data (e.g., previous speed values, topological features, etc.) up to a certain time t, what will be the average speed on a given set of road segments at time t + T , where T is the prediction horizon.
More specifically, we focus on short term prediction (i.e., a maximum T of one hour) for a subset of road segments in the road network of Lyon, France, by exploiting the past history of observed traffic speeds from a real dataset
Origine | Fichiers produits par l'(les) auteur(s) |
---|