Abstract: We present BusTr, a machine-learned model for translating road tra c forecasts into predictions of bus delays, used by Google Maps to serve the majority of the world’s public transit systems where no o cial real-time bus tracking is provided. We demonstrate that our neural sequence model improves over DeepTTE, the state-of-the-art baseline, both in performance (−30% MAPE) and training stability. We also demonstrate signi cant generalization gains over simpler models, evaluated on longitudinal data to cope with a constantly evolving world.
BusTr: Predicting Bus Travel Times from Real-Time Traffic
Richard Barnes, Alex Fabrikant, Senaka Buthpitiya, Andrew Tomkins, James Cook, Fangzhou Xu
KDD '20 Proceedings
August 23, 2020