Artificial Neural Networks Applied to Taxi Destination Prediction
This work addresses a domain-specific problem for urban transportation and logistics, with incremental improvements in method application.
The authors tackled the problem of predicting taxi destinations from partial GPS trajectories and metadata, achieving first place out of 381 teams in the ECML/PKDD challenge.
We describe our first-place solution to the ECML/PKDD discovery challenge on taxi destination prediction. The task consisted in predicting the destination of a taxi based on the beginning of its trajectory, represented as a variable-length sequence of GPS points, and diverse associated meta-information, such as the departure time, the driver id and client information. Contrary to most published competitor approaches, we used an almost fully automated approach based on neural networks and we ranked first out of 381 teams. The architectures we tried use multi-layer perceptrons, bidirectional recurrent neural networks and models inspired from recently introduced memory networks. Our approach could easily be adapted to other applications in which the goal is to predict a fixed-length output from a variable-length sequence.