{"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/3d-graph-convolutional-networks-with-temporal","title":"3D Graph Convolutional Networks with Temporal Graphs: A Spatial Information Free Framework For Traffic Forecasting","arxiv_id":"1903.00919","date":"2019-03-03","proceeding":null,"authors":["Bing Yu","Mengzhang Li","Jiyong Zhang","Zhanxing Zhu"],"abstract":"Spatio-temporal prediction plays an important role in many application areas\nespecially in traffic domain. However, due to complicated spatio-temporal\ndependency and high non-linear dynamics in road networks, traffic prediction\ntask is still challenging. Existing works either exhibit heavy training cost or\nfail to accurately capture the spatio-temporal patterns, also ignore the\ncorrelation between distant roads that share the similar patterns. In this\npaper, we propose a novel deep learning framework to overcome these issues: 3D\nTemporal Graph Convolutional Networks (3D-TGCN). Two novel components of our\nmodel are introduced. (1) Instead of constructing the road graph based on\nspatial information, we learn it by comparing the similarity between time\nseries for each road, thus providing a spatial information free framework. (2)\nWe propose an original 3D graph convolution model to model the spatio-temporal\ndata more accurately. Empirical results show that 3D-TGCN could outperform\nstate-of-the-art baselines.","url_abs":"http://arxiv.org/abs/1903.00919v1","url_pdf":"http://arxiv.org/pdf/1903.00919v1.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":[],"tasks":[{"task_slug":"4d-spatio-temporal-semantic-segmentation","task_name":"4D Spatio Temporal Semantic Segmentation"},{"task_slug":"time-series-1","task_name":"Time Series"},{"task_slug":"time-series","task_name":"Time Series Analysis"},{"task_slug":"traffic-prediction","task_name":"Traffic Prediction"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"graph-convolutional-networks","method_name":"Graph Convolutional Networks"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/traffic-prediction-on-pems-m","task":"Traffic Prediction","dataset":"PeMS-M","model":"3D-TGCN","rank_in_archive_order":2,"of":5,"metrics":{"MAE (60 min)":"3.65"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.00919","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}