Enhancing spatiotemporal prediction through the integration of Mamba state space models and Diffusion Transformers

doi:10.1016/j.knosys.2025.113347

	Enhancing spatiotemporal prediction through the integration of Mamba state space models and Diffusion Transformers
	Zeng, Hansheng 1; Li, Yuqi 2; Niu, Ruize 3; Yang, Chuanguang 2; Wen, Shiping 4
	2025-05-12
发表期刊	KNOWLEDGE-BASED SYSTEMS
ISSN	0950-7051
卷号	316 页码:11
摘要	This paper presents an advanced architecture for spatiotemporal prediction MAD, integrating Mamba modules with Diffusion Transformers for efficient spatiotemporal modeling. The model consists of three phases: encoding, reconstruction, and prediction. Initially, the encoder transforms raw spatiotemporal data into compact latent embeddings. In the reconstruction phase, the Mamba module processes these embeddings through normalization and bidirectional state space models, generating reconstructed representations which are then decoded to restore the input data. The prediction phase utilizes the Diffusion Transformer to model spatiotemporal features, incorporating time embeddings and leveraging self-attention mechanisms to capture complex spatiotemporal dependencies. Finally, the model jointly trains the reconstruction and prediction paths to achieve high-precision spatiotemporal forecasts. Experimental results demonstrate the model's superior performance across various spatiotemporal prediction tasks, validating its effectiveness and robustness. Our codes are available at https://github.com/Hanson1331/KBS-MAD.
关键词	Deep learning Spatio-temporal prediction Mamba Diffusion
DOI	10.1016/j.knosys.2025.113347
收录类别	SCI
语种	英语
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:001460554900001
出版者	ELSEVIER
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40659
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Wen, Shiping
作者单位	1.Univ Hong Kong, Hong Kong, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China 3.Sun Yat Sen Univ, Guangzhou, Peoples R China 4.Univ Technol, Australian AI Inst, Fac Engn & Informat Technol, Ultimo, Australia
推荐引用方式 GB/T 7714	Zeng, Hansheng,Li, Yuqi,Niu, Ruize,et al. Enhancing spatiotemporal prediction through the integration of Mamba state space models and Diffusion Transformers[J]. KNOWLEDGE-BASED SYSTEMS,2025,316:11.
APA	Zeng, Hansheng,Li, Yuqi,Niu, Ruize,Yang, Chuanguang,&Wen, Shiping.(2025).Enhancing spatiotemporal prediction through the integration of Mamba state space models and Diffusion Transformers.KNOWLEDGE-BASED SYSTEMS,316,11.
MLA	Zeng, Hansheng,et al."Enhancing spatiotemporal prediction through the integration of Mamba state space models and Diffusion Transformers".KNOWLEDGE-BASED SYSTEMS 316(2025):11.