ARMemNet: Autoregressive memory networks for multivariate time series forecasting

Jinuk Park; Chanhee Park; Hongchan Roh; Sanghyun Park

doi:10.1145/3412841.3442108

ARMemNet: Autoregressive memory networks for multivariate time series forecasting

Jinuk Park, Chanhee Park, Hongchan Roh, Sanghyun Park

College of Computing

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Recently, several studies show the powerful capability of neural networks to capture non-linear features from time series which have multiple seasonal patterns. However, existing methods rely on convolution kernels implicitly, hence neglect to capture strong long-term patterns and lack interpretability. In this paper, we propose a memory-augmented neural network named AutoRegressive Memory Network (ARMemNet) for multivariate time series forecasting. ARMemNet utilizes memory components to explicitly encode intense long-term patterns. Furthermore, each encoder is designed to leverage inherently essential autoregressive property to represent short-term patterns. In experiments on real-world dataset, ARMemNet outperforms existing baselines and validates effectiveness of memory components for complex seasonality which is prevalent in time series datasets.

Original language	English
Title of host publication	Proceedings of the 36th Annual ACM Symposium on Applied Computing, SAC 2021
Publisher	Association for Computing Machinery
Pages	1094-1097
Number of pages	4
ISBN (Electronic)	9781450381048
DOIs	https://doi.org/10.1145/3412841.3442108
Publication status	Published - 2021 Mar 22
Event	36th Annual ACM Symposium on Applied Computing, SAC 2021 - Virtual, Online, Korea, Republic of Duration: 2021 Mar 22 → 2021 Mar 26

Publication series

Name	Proceedings of the ACM Symposium on Applied Computing

Conference

Conference	36th Annual ACM Symposium on Applied Computing, SAC 2021
Country/Territory	Korea, Republic of
City	Virtual, Online
Period	21/3/22 → 21/3/26

Bibliographical note

Publisher Copyright:
© 2021 Owner/Author.

All Science Journal Classification (ASJC) codes

Software

Access to Document

10.1145/3412841.3442108

Cite this

@inproceedings{492dddc88de840dfb2f54933ea4a9c13,

title = "ARMemNet: Autoregressive memory networks for multivariate time series forecasting",

abstract = "Recently, several studies show the powerful capability of neural networks to capture non-linear features from time series which have multiple seasonal patterns. However, existing methods rely on convolution kernels implicitly, hence neglect to capture strong long-term patterns and lack interpretability. In this paper, we propose a memory-augmented neural network named AutoRegressive Memory Network (ARMemNet) for multivariate time series forecasting. ARMemNet utilizes memory components to explicitly encode intense long-term patterns. Furthermore, each encoder is designed to leverage inherently essential autoregressive property to represent short-term patterns. In experiments on real-world dataset, ARMemNet outperforms existing baselines and validates effectiveness of memory components for complex seasonality which is prevalent in time series datasets.",

author = "Jinuk Park and Chanhee Park and Hongchan Roh and Sanghyun Park",

note = "Publisher Copyright: {\textcopyright} 2021 Owner/Author.; 36th Annual ACM Symposium on Applied Computing, SAC 2021 ; Conference date: 22-03-2021 Through 26-03-2021",

year = "2021",

month = mar,

day = "22",

doi = "10.1145/3412841.3442108",

language = "English",

series = "Proceedings of the ACM Symposium on Applied Computing",

publisher = "Association for Computing Machinery",

pages = "1094--1097",

booktitle = "Proceedings of the 36th Annual ACM Symposium on Applied Computing, SAC 2021",

}

Park, J, Park, C, Roh, H & Park, S 2021, ARMemNet: Autoregressive memory networks for multivariate time series forecasting. in Proceedings of the 36th Annual ACM Symposium on Applied Computing, SAC 2021. Proceedings of the ACM Symposium on Applied Computing, Association for Computing Machinery, pp. 1094-1097, 36th Annual ACM Symposium on Applied Computing, SAC 2021, Virtual, Online, Korea, Republic of, 21/3/22. https://doi.org/10.1145/3412841.3442108

ARMemNet: Autoregressive memory networks for multivariate time series forecasting. / Park, Jinuk; Park, Chanhee; Roh, Hongchan et al.
Proceedings of the 36th Annual ACM Symposium on Applied Computing, SAC 2021. Association for Computing Machinery, 2021. p. 1094-1097 (Proceedings of the ACM Symposium on Applied Computing).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - ARMemNet

T2 - 36th Annual ACM Symposium on Applied Computing, SAC 2021

AU - Park, Jinuk

AU - Park, Chanhee

AU - Roh, Hongchan

AU - Park, Sanghyun

PY - 2021/3/22

Y1 - 2021/3/22

N2 - Recently, several studies show the powerful capability of neural networks to capture non-linear features from time series which have multiple seasonal patterns. However, existing methods rely on convolution kernels implicitly, hence neglect to capture strong long-term patterns and lack interpretability. In this paper, we propose a memory-augmented neural network named AutoRegressive Memory Network (ARMemNet) for multivariate time series forecasting. ARMemNet utilizes memory components to explicitly encode intense long-term patterns. Furthermore, each encoder is designed to leverage inherently essential autoregressive property to represent short-term patterns. In experiments on real-world dataset, ARMemNet outperforms existing baselines and validates effectiveness of memory components for complex seasonality which is prevalent in time series datasets.

AB - Recently, several studies show the powerful capability of neural networks to capture non-linear features from time series which have multiple seasonal patterns. However, existing methods rely on convolution kernels implicitly, hence neglect to capture strong long-term patterns and lack interpretability. In this paper, we propose a memory-augmented neural network named AutoRegressive Memory Network (ARMemNet) for multivariate time series forecasting. ARMemNet utilizes memory components to explicitly encode intense long-term patterns. Furthermore, each encoder is designed to leverage inherently essential autoregressive property to represent short-term patterns. In experiments on real-world dataset, ARMemNet outperforms existing baselines and validates effectiveness of memory components for complex seasonality which is prevalent in time series datasets.

UR - http://www.scopus.com/inward/record.url?scp=85104984549&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85104984549&partnerID=8YFLogxK

U2 - 10.1145/3412841.3442108

DO - 10.1145/3412841.3442108

M3 - Conference contribution

AN - SCOPUS:85104984549

T3 - Proceedings of the ACM Symposium on Applied Computing

SP - 1094

EP - 1097

BT - Proceedings of the 36th Annual ACM Symposium on Applied Computing, SAC 2021

PB - Association for Computing Machinery

Y2 - 22 March 2021 through 26 March 2021

ER -

ARMemNet: Autoregressive memory networks for multivariate time series forecasting

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this