Cluster-aware attentive convolutional recurrent network for multivariate time-series forecasting

Bai, S; Zhang, Q; He, H; Hu, L; Wang, S; Niu, Z

Cluster-aware attentive convolutional recurrent network for multivariate time-series forecasting

Bai, S Zhang, Q He, H Hu, L Wang, S

Niu, Z

Permalink

Publisher:: ELSEVIER
Publication Type:: Journal Article
Citation:: Neurocomputing, 2023, 558
Issue Date:: 2023-11-14

Closed Access

	Filename	Description	Size
	65. NeuroComputing. MTS. 2023. Simeng.pdf		1.82 MB		View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Bai, S
dc.contributor.author	Zhang, Q
dc.contributor.author	He, H
dc.contributor.author	Hu, L
dc.contributor.author	Wang, S https://orcid.org/0000-0003-1133-9379
dc.contributor.author	Niu, Z
dc.date.accessioned	2023-10-31T23:34:29Z
dc.date.available	2023-10-31T23:34:29Z
dc.date.issued	2023-11-14
dc.identifier.citation	Neurocomputing, 2023, 558
dc.identifier.issn	0925-2312
dc.identifier.issn	1872-8286
dc.identifier.uri	http://hdl.handle.net/10453/172912
dc.description.abstract	Multivariate time-series (MTS) forecasting plays a crucial role in various real-world applications, but the complex dependencies between time-series variables (i.e., inter-series dependencies) make this task extremely challenging. While most existing studies focus on modeling intra-series (temporal) dependencies by capturing long- and short-term patterns, they fail to explore and exploit the inter-series dependencies to enhance MTS forecasting. In this paper, we propose a Cluster-aware Attentive Convolutional Recurrent Network (CACRN) to capture both inter-series and intra-series dependencies in MTS data. Specifically, CACRN first introduces a cluster-aware variable representation module that separates irrelevant variables and captures the interaction between relevant variables to learn cluster-aware variable representations. Then, CACRN feeds these representations into parallel convolutional recurrent neural networks (CRNNs) to capture the short- and long-term temporal dependencies in a cluster-wise manner. Next, a cluster-aware attention mechanism is introduced to attend to temporal information in each cluster and co-attend all cluster information jointly to capture intra-cluster and inter-cluster dependencies for the downstream forecasting task. Our extensive experiments on six real-world datasets demonstrate that CACRN is effective and outperforms representative and state-of-the-art baselines. Our proposed method is suitable for a wide range of real-world data collections, especially those with clear dependencies of variables.
dc.language	English
dc.publisher	ELSEVIER
dc.relation.ispartof	Neurocomputing
dc.relation.isbasedon	10.1016/j.neucom.2023.126701
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	08 Information and Computing Sciences, 09 Engineering, 17 Psychology and Cognitive Sciences
dc.subject.classification	Artificial Intelligence & Image Processing
dc.subject.classification	40 Engineering
dc.subject.classification	46 Information and computing sciences
dc.subject.classification	52 Psychology
dc.title	Cluster-aware attentive convolutional recurrent network for multivariate time-series forecasting
dc.type	Journal Article
utslib.citation.volume	558
utslib.for	08 Information and Computing Sciences
utslib.for	09 Engineering
utslib.for	17 Psychology and Cognitive Sciences
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
utslib.copyright.status	closed_access	*
dc.date.updated	2023-10-31T23:34:26Z
pubs.publication-status	Accepted
pubs.volume	558

Abstract:

Multivariate time-series (MTS) forecasting plays a crucial role in various real-world applications, but the complex dependencies between time-series variables (i.e., inter-series dependencies) make this task extremely challenging. While most existing studies focus on modeling intra-series (temporal) dependencies by capturing long- and short-term patterns, they fail to explore and exploit the inter-series dependencies to enhance MTS forecasting. In this paper, we propose a Cluster-aware Attentive Convolutional Recurrent Network (CACRN) to capture both inter-series and intra-series dependencies in MTS data. Specifically, CACRN first introduces a cluster-aware variable representation module that separates irrelevant variables and captures the interaction between relevant variables to learn cluster-aware variable representations. Then, CACRN feeds these representations into parallel convolutional recurrent neural networks (CRNNs) to capture the short- and long-term temporal dependencies in a cluster-wise manner. Next, a cluster-aware attention mechanism is introduced to attend to temporal information in each cluster and co-attend all cluster information jointly to capture intra-cluster and inter-cluster dependencies for the downstream forecasting task. Our extensive experiments on six real-world datasets demonstrate that CACRN is effective and outperforms representative and state-of-the-art baselines. Our proposed method is suitable for a wide range of real-world data collections, especially those with clear dependencies of variables.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/172912