Meta-MSNet: Meta-Learning Based Multi-Source Data Fusion for Traffic Flow Prediction

doi:10.1109/LSP.2020.3037527

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 先进时空数据分析与学习

	Meta-MSNet: Meta-Learning Based Multi-Source Data Fusion for Traffic Flow Prediction
	Fang, Shen1,2 ; Pan, Xianbing 3; Xiang, Shiming1,2 ; Pan, Chunhong1
发表期刊	IEEE SIGNAL PROCESSING LETTERS
ISSN	1070-9908
	2021
卷号	28 页码:6-10
通讯作者	Xiang, Shiming(smxiang@nlpr.ia.ac.cn)
摘要	Traffic flow prediction is a challenging task while most existing works are faced with two main problems in extracting complicated intrinsic and extrinsic features. In terms of intrinsic features, current methods don't fully exploit different functions of short-term neighboring and long-term periodic temporal patterns. As for extrinsic features, recent works mainly employ hand-crafted fusion strategies to integrate external factors but remain generalization issues. To solve these problems, we propose a meta-learning based multi-source spatio-temporal network (Meta-MSNet). The Meta-MSNet is designed with an encoder-decoder structure. The encoder captures neighboring temporal dependencies while the decoder extracts periodic features. Furthermore, two meta-learning based fusion modules are designed to integrate multi-source external data both on temporal and spatial dimensions. Experiments on three real-world traffic datasets have verified the superiority of the proposed model.
关键词	Data fusion deep learning graph convolution meta-learning traffic flow prediction traffic network
DOI	10.1109/LSP.2020.3037527
收录类别	SCI
语种	英语
资助项目	Major Project for New Generation of AI[2018AAA0100400] ; NationalNatural Science Foundation ofChina[91646207] ; NationalNatural Science Foundation ofChina[62076242] ; NationalNatural Science Foundation ofChina[61773377] ; NationalNatural Science Foundation ofChina[61976208]
项目资助者	Major Project for New Generation of AI ; NationalNatural Science Foundation ofChina
WOS研究方向	Engineering
WOS类目	Engineering, Electrical & Electronic
WOS记录号	WOS:000608679700002
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
七大方向——子方向分类	机器学习
引用统计	被引频次：18[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/42591
专题	多模态人工智能系统全国重点实验室_先进时空数据分析与学习中国科学院自动化研究所
通讯作者	Xiang, Shiming
作者单位	1.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 101408, Peoples R China 3.Chongqing Univ Posts & Telecommun, Coll Mobile Telecommun, Chongqing 400065, Peoples R China
第一作者单位	模式识别国家重点实验室
通讯作者单位	模式识别国家重点实验室
推荐引用方式 GB/T 7714	Fang, Shen,Pan, Xianbing,Xiang, Shiming,et al. Meta-MSNet: Meta-Learning Based Multi-Source Data Fusion for Traffic Flow Prediction[J]. IEEE SIGNAL PROCESSING LETTERS,2021,28:6-10.
APA	Fang, Shen,Pan, Xianbing,Xiang, Shiming,&Pan, Chunhong.(2021).Meta-MSNet: Meta-Learning Based Multi-Source Data Fusion for Traffic Flow Prediction.IEEE SIGNAL PROCESSING LETTERS,28,6-10.
MLA	Fang, Shen,et al."Meta-MSNet: Meta-Learning Based Multi-Source Data Fusion for Traffic Flow Prediction".IEEE SIGNAL PROCESSING LETTERS 28(2021):6-10.