Building on its success at NeurIPS 2019, the Traffic4cast competition is going into its second year offering new challenges and opportunities. This year’s dataset will be derived from an order of magnitude more data. We are collecting data from 10 cities of different size, geography, culture, and economy. The core challenge will be to predict short-term large-scale traffic states in all the cities. Participants can investigate differences and similarities in traffic patterns between the cities, and explore master models trained on multiple cities. The dataset will be augmented by new static and dynamic features, such as street maps properties, points of interest, weather, air pollution, and special events. The bonus challenges invite participants to explore the effects of these additional features.
Our competition has completed. You can find a description of our data, of winning approaches and further learnings in the reference below.
A detaild desciption can be found here and details on the code, submission format, metric and data here.
A summary of our competition - including the data provided, winning approaches and what we learned - can be found in the following reference. Please also cite it if you refer to our traffic4cast competition.
@InProceedings{pmlr-v133-kopp21a,
title = {Traffic4cast at NeurIPS 2020 - yet more on the unreasonable effectiveness of gridded geo-spatial processes},
author = {Kopp, Michael and Kreil, David and Neun, Moritz and Jonietz, David and Martin, Henry and Herruzo, Pedro and Gruca, Aleksandra and Soleymani, Ali and Wu, Fanyou and Liu, Yang and Xu, Jingwei and Zhang, Jianjin and Santokhi, Jay and Bojesomo, Alabi and Marzouqi, Hasan Al and Liatsis, Panos and Kwok, Pak Hay and Qi, Qi and Hochreiter, Sepp},
booktitle = {Proceedings of the NeurIPS 2020 Competition and Demonstration Track},
pages = {325--343},
year = {2021},
editor = {Escalante, Hugo Jair and Hofmann, Katja},
volume = {133},
series = {Proceedings of Machine Learning Research},
month = {06--12 Dec},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v133/kopp21a/kopp21a.pdf},
url = {https://proceedings.mlr.press/v133/kopp21a.html}
}