Spatial transition learning on road networks with deep probabilistic models

In this paper, we study the problem of predicting the most likely traveling route on the road network between two given locations by considering the real-time traffic. We present a deep probabilistic model-<monospce>DeepST</monospce>-which unifies three key explanatory factors, the past traveled route, the impact of destination and real-time traffic for the route decision. <monospce>DeepST</monospce> explains the generation of next route by conditioning on the representations of the three explanatory factors. To enable effectively sharing the statistical strength, we propose to learn representations of K-destination proxies with an adjoint generative model. To incorporate the impact of real-time traffic, we introduce a high dimensional latent variable as its representation whose posterior distribution can then be inferred from observations. An efficient inference method is developed within the Variational Auto-Encoders framework to scale <monospce>DeepST</monospce> to large-scale datasets. We conduct experiments on two real-world large-scale trajectory datasets to demonstrate the superiority of <monospce>DeepST</monospce> over the existing methods on two tasks: the most likely route prediction and route recovery from sparse trajectories. In particular, on one public large-scale trajectory dataset, <monospce>DeepST</monospce> surpasses the best competing method by almost 50% on the most likely route prediction task and up to 15% on the route recovery task in terms of accuracy.