{"type":"ScholarlyArticle","title":"StreamMTS: Towards Streaming Multivariate Time Series Forecasting","authors":[{"name":"Binwu Wang","slug":"binwu-wang","given":"Binwu","family":"Wang","affiliation":""},{"name":"Jiaming Ma","slug":"jiaming-ma","given":"Jiaming","family":"Ma","affiliation":""},{"name":"Yudong Zhang","slug":"yudong-zhang","given":"Yudong","family":"Zhang","affiliation":""},{"name":"Pengkun Wang","slug":"pengkun-wang","given":"Pengkun","family":"Wang","affiliation":""},{"name":"Zhengyang Zhou","slug":"zhengyang-zhou","given":"Zhengyang","family":"Zhou","affiliation":""},{"name":"Xu Wang","slug":"xu-wang","given":"Xu","family":"Wang","affiliation":""},{"name":"Yang Wang","slug":"yang-wang","given":"Yang","family":"Wang","affiliation":""}],"abstract":"Current mainstream research in multivariate time series (MTS) forecasting often assumes that all data are static. However, real-world MTS data typically arrives continuously in a streaming manner, which we refer to as streaming MTS. The statistical characteristics and spatiotemporal graph topology of these data evolve over time, presenting two key challenges: the model's ability to adapt to new data distributions and the enhancement of cross-domain generalization capabilities. In this paper, we propose a streaming MTS prediction framework. We begin by designing a lightweight spatiotemporal causal learning model that captures generalizable causal spatiotemporal features from a decoupling perspective. Next, we introduce a framework to enhance the model's streaming learning capability, leveraging the adaptability of continual learning while strengthening cross-domain representation abilities. Specifically, we reformulate continual learning as a multi-task learning problem and present a multi-task optimization algorithm that identifies a set of Pareto-optimal solutions to address the inherent stability-plasticity dilemma in continual learning. Finally, we propose a topology-aware feature propagation strategy that disseminates well-trained node embedding features to unseen graph structures, thereby improving the model's cross-domain generalization. Results on real-world datasets demonstrate that our model achieves superior forecasting performance while substantially improving computational efficiency and memory usage.","doi":"10.24963/ijcai.2026/340","url":"https://ijcai.org/proceedings/2026/340","pdf":"https://ijcai.org/proceedings/2026/0340.pdf","markdown":"https://ijcai.org/proceedings/2026/340.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/340","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Data Mining","track":"Main Track","number":340,"pages":{"first":3061,"last":3069},"keywords":["Data Mining: Mining spatial and/or temporal data"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}