{"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/spatio-temporal-graph-mixformer-for-traffic","title":"Spatio-Temporal Graph Mixformer for Traffic Forecasting","arxiv_id":null,"date":"2023-10-15","proceeding":"Expert Systems with Applications 2023 10","authors":["Mourad Lablack","Yanming Shen"],"abstract":"Traffic forecasting is of great importance for intelligent transportation systems (ITS). Because of the intricacy implied in traffic behavior and the non-Euclidean nature of traffic data, it is challenging to give an accurate traffic prediction. Despite that previous studies considered the relationship between different nodes, the majority have relied on a static representation and failed to capture the dynamic node interactions over time. Additionally, prior studies employed RNN-based models to capture the temporal dependency. While RNNs are a popular choice for forecasting problems, they tend to be memory hungry and slow to train. Furthermore, recent studies start utilizing similarity algorithms to better express the implication of a node over the other. However, to our knowledge, none have explored the contribution of node $𝑖$’s past, over the future state of node $𝑗$. In this paper, we propose a Spatio-Temporal Graph Mixformer (STGM) network, a highly optimized model with low memory footprint. We address the aforementioned limits by utilizing a novel attention mechanism to capture the correlation between temporal and spatial dependencies. Specifically, we use convolution layers with a variable field of view for each head to capture long–short term temporal dependency. Additionally, we train an estimator model that express the contribution of a node over the desired prediction. The estimation is fed alongside a distance matrix to the attention mechanism. Meanwhile, we use a gated mechanism and a mixer layer to further select and incorporate the different perspectives. Extensive experiments show that the proposed model enjoys a performance gain compared to the baselines while maintaining the lowest parameter counts.","url_abs":"https://doi.org/10.1016/j.eswa.2023.120281","url_pdf":"https://doi.org/10.1016/j.eswa.2023.120281","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":"spatio-temporal-graph-mixformer-for-traffic","repo_url":"https://github.com/Mouradost/STGM","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"traffic-prediction","task_name":"Traffic Prediction"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dilated-convolution","method_name":"Dilated Convolution"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"graph-transformer","method_name":"Graph Transformer"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"mixer-layer","method_name":"Mixer Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":null,"method_name":"None"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/traffic-prediction-on-metr-la","task":"Traffic Prediction","dataset":"METR-LA","model":"STGM","rank_in_archive_order":3,"of":20,"metrics":{"12 steps MAE":"3.229","12 steps MAPE":"9.39","12 steps RMSE":"7.099","MAE @ 12 step":"3.229"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pems-bay","task":"Traffic Prediction","dataset":"PEMS-BAY","model":"STGM","rank_in_archive_order":6,"of":16,"metrics":{"MAE @ 12 step":"1.857","RMSE ":"4.369"},"uses_additional_data":false},{"leaderboard":"/sota/traffic-prediction-on-pemsd7-m","task":"Traffic Prediction","dataset":"PeMSD7(M)","model":"STGM","rank_in_archive_order":7,"of":7,"metrics":{"12 steps MAE":"3.002","12 steps MAPE":"8.01","12 steps RMSE":"6.331"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}