{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/graph-neural-controlled-differential","title":"Graph Neural Controlled Differential Equations for Traffic Forecasting","arxiv_id":"2112.03558","date":"2021-12-07","proceeding":null,"authors":["Jeongwhan Choi","Hwangyong Choi","Jeehyun Hwang","Noseong Park"],"abstract":"Traffic forecasting is one of the most popular spatio-temporal tasks in the field of machine learning. A prevalent approach in the field is to combine graph convolutional networks and recurrent neural networks for the spatio-temporal processing. There has been fierce competition and many novel methods have been proposed. In this paper, we present the method of spatio-temporal graph neural controlled differential equation (STG-NCDE). Neural controlled differential equations (NCDEs) are a breakthrough concept for processing sequential data. We extend the concept and design two NCDEs: one for the temporal processing and the other for the spatial processing. After that, we combine them into a single framework. We conduct experiments with 6 benchmark datasets and 20 baselines. STG-NCDE shows the best accuracy in all cases, outperforming all those 20 baselines by non-trivial margins.","url_abs":"https://arxiv.org/abs/2112.03558v1","url_pdf":"https://arxiv.org/pdf/2112.03558v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"graph-neural-controlled-differential","repo_url":"https://github.com/jeongwhanchoi/STG-NCDE","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"spatio-temporal-forecasting","task_name":"Spatio-Temporal Forecasting"},{"task_slug":"time-series-forecasting","task_name":"Time Series Forecasting"},{"task_slug":"traffic-prediction","task_name":"Traffic Prediction"},{"task_slug":"weather-forecasting","task_name":"Weather Forecasting"}],"methods":[{"method_slug":"gcn","method_name":"GCN"},{"method_slug":"node","method_name":"NODE"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/traffic-prediction-on-pemsd3","task":"Traffic Prediction","dataset":"PeMSD3","model":"STG-NCDE","rank_in_archive_order":6,"of":6,"metrics":{"12 steps MAE":"15.57","12 steps MAPE":"15.06","12 steps RMSE":"27.09"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pemsd4","task":"Traffic Prediction","dataset":"PeMSD4","model":"STG-NCDE","rank_in_archive_order":13,"of":13,"metrics":{"12 steps MAE":"19.21","12 steps MAPE":"12.76","12 steps RMSE":"31.09"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pemsd7","task":"Traffic Prediction","dataset":"PeMSD7","model":"STG-NCDE","rank_in_archive_order":8,"of":8,"metrics":{"12 steps MAE":"20.53","12 steps MAPE":"8.8","12 steps RMSE":"33.84"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pemsd7-l","task":"Traffic Prediction","dataset":"PeMSD7(L)","model":"STG-NCDE","rank_in_archive_order":6,"of":6,"metrics":{"12 steps MAE":"2.87","12 steps MAPE":"7.31","12 steps RMSE":"5.76"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pemsd7-m","task":"Traffic Prediction","dataset":"PeMSD7(M)","model":"STG-NCDE","rank_in_archive_order":6,"of":7,"metrics":{"12 steps MAE":"2.68","12 steps MAPE":"6.76","12 steps RMSE":"5.39"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pemsd8","task":"Traffic Prediction","dataset":"PeMSD8","model":"STG-NCDE","rank_in_archive_order":13,"of":13,"metrics":{"12 steps MAE":"15.45","12 steps MAPE":"9.92","12 steps RMSE":"24.81"},"uses_additional_data":false},{"leaderboard":"/sota/weather-forecasting-on-noaa-atmospheric","task":"Weather Forecasting","dataset":"NOAA Atmospheric Temperature Dataset","model":"STG-NCDE","rank_in_archive_order":5,"of":5,"metrics":{"MAE (t+1)":"0.3582 ± 0.0616","MAE (t+10)":"1.4095 ± 0.1836"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2112.03558","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}