Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network

Hyung Jun Moon; Seok Jun Bu; Sung Bae Cho

doi:10.1109/ICDMW51313.2020.00110

Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network

Hyung Jun Moon, Seok Jun Bu, Sung Bae Cho

College of Computing

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

2 Citations (Scopus)

Abstract

In the time-series models for predicting residential energy consumption, the energy properties collected through multiple sensors usually include irregular and seasonal factors. The irregular pattern resulting from them is called peak demand, which is a major cause of performance degradation. In order to enhance the performance, we propose a convolutional-recurrent triplet network to learn and detect the demand peaks. The proposed model generates the latent space for demand peaks from data, which is transferred into convolutional neural network-long short-term memory (CNN-LSTM) to finally predict the future power demand. Experiments with the dataset of UCI household power consumption composed of a total of 2,075,259 time-series data show that the proposed model reduces the error by 23.63% and outperforms the state-of-the-art deep learning models including the CNN-LSTM. Especially, the proposed model improves the prediction performance by modeling the distribution of demand peaks in Euclidean space.

Original language	English
Title of host publication	Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020
Editors	Giuseppe Di Fatta, Victor Sheng, Alfredo Cuzzocrea, Carlo Zaniolo, Xindong Wu
Publisher	IEEE Computer Society
Pages	757-761
Number of pages	5
ISBN (Electronic)	9781728190129
DOIs	https://doi.org/10.1109/ICDMW51313.2020.00110
Publication status	Published - 2020 Nov
Event	20th IEEE International Conference on Data Mining Workshops, ICDMW 2020 - Virtual, Sorrento, Italy Duration: 2020 Nov 17 → 2020 Nov 20

Publication series

Name	IEEE International Conference on Data Mining Workshops, ICDMW
Volume	2020-November
ISSN (Print)	2375-9232
ISSN (Electronic)	2375-9259

Conference

Conference	20th IEEE International Conference on Data Mining Workshops, ICDMW 2020
Country/Territory	Italy
City	Virtual, Sorrento
Period	20/11/17 → 20/11/20

Bibliographical note

Funding Information:
ACKNOWLEDGEMENTS This work was partly supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) (No. 2020-0-01361, Artificial Intelligence Graduate School Program (Yonsei University)) and Korea Electric Power Corporation (Grant number: R18XA05).

Publisher Copyright:
© 2020 IEEE.

All Science Journal Classification (ASJC) codes

Computer Science Applications
Software

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/ICDMW51313.2020.00110

Cite this

Moon, H. J., Bu, S. J., & Cho, S. B. (2020). Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network. In G. Di Fatta, V. Sheng, A. Cuzzocrea, C. Zaniolo, & X. Wu (Eds.), Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020 (pp. 757-761). [9346355] (IEEE International Conference on Data Mining Workshops, ICDMW; Vol. 2020-November). IEEE Computer Society. https://doi.org/10.1109/ICDMW51313.2020.00110

Moon, Hyung Jun ; Bu, Seok Jun ; Cho, Sung Bae. / Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network. Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020. editor / Giuseppe Di Fatta ; Victor Sheng ; Alfredo Cuzzocrea ; Carlo Zaniolo ; Xindong Wu. IEEE Computer Society, 2020. pp. 757-761 (IEEE International Conference on Data Mining Workshops, ICDMW).

@inproceedings{523dd5ef0a9b45e88e56f91dff657ebc,

title = "Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network",

abstract = "In the time-series models for predicting residential energy consumption, the energy properties collected through multiple sensors usually include irregular and seasonal factors. The irregular pattern resulting from them is called peak demand, which is a major cause of performance degradation. In order to enhance the performance, we propose a convolutional-recurrent triplet network to learn and detect the demand peaks. The proposed model generates the latent space for demand peaks from data, which is transferred into convolutional neural network-long short-term memory (CNN-LSTM) to finally predict the future power demand. Experiments with the dataset of UCI household power consumption composed of a total of 2,075,259 time-series data show that the proposed model reduces the error by 23.63% and outperforms the state-of-the-art deep learning models including the CNN-LSTM. Especially, the proposed model improves the prediction performance by modeling the distribution of demand peaks in Euclidean space.",

author = "Moon, {Hyung Jun} and Bu, {Seok Jun} and Cho, {Sung Bae}",

note = "Funding Information: ACKNOWLEDGEMENTS This work was partly supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) (No. 2020-0-01361, Artificial Intelligence Graduate School Program (Yonsei University)) and Korea Electric Power Corporation (Grant number: R18XA05). Publisher Copyright: {\textcopyright} 2020 IEEE.; 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020 ; Conference date: 17-11-2020 Through 20-11-2020",

year = "2020",

month = nov,

doi = "10.1109/ICDMW51313.2020.00110",

language = "English",

series = "IEEE International Conference on Data Mining Workshops, ICDMW",

publisher = "IEEE Computer Society",

pages = "757--761",

editor = "{Di Fatta}, Giuseppe and Victor Sheng and Alfredo Cuzzocrea and Carlo Zaniolo and Xindong Wu",

booktitle = "Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020",

address = "United States",

}

Moon, HJ, Bu, SJ & Cho, SB 2020, Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network. in G Di Fatta, V Sheng, A Cuzzocrea, C Zaniolo & X Wu (eds), Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020., 9346355, IEEE International Conference on Data Mining Workshops, ICDMW, vol. 2020-November, IEEE Computer Society, pp. 757-761, 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020, Virtual, Sorrento, Italy, 20/11/17. https://doi.org/10.1109/ICDMW51313.2020.00110

Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network. / Moon, Hyung Jun; Bu, Seok Jun; Cho, Sung Bae.
Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020. ed. / Giuseppe Di Fatta; Victor Sheng; Alfredo Cuzzocrea; Carlo Zaniolo; Xindong Wu. IEEE Computer Society, 2020. p. 757-761 9346355 (IEEE International Conference on Data Mining Workshops, ICDMW; Vol. 2020-November).

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

TY - GEN

T1 - Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network

AU - Moon, Hyung Jun

AU - Bu, Seok Jun

AU - Cho, Sung Bae

N1 - Funding Information: ACKNOWLEDGEMENTS This work was partly supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) (No. 2020-0-01361, Artificial Intelligence Graduate School Program (Yonsei University)) and Korea Electric Power Corporation (Grant number: R18XA05). Publisher Copyright: © 2020 IEEE.

PY - 2020/11

Y1 - 2020/11

N2 - In the time-series models for predicting residential energy consumption, the energy properties collected through multiple sensors usually include irregular and seasonal factors. The irregular pattern resulting from them is called peak demand, which is a major cause of performance degradation. In order to enhance the performance, we propose a convolutional-recurrent triplet network to learn and detect the demand peaks. The proposed model generates the latent space for demand peaks from data, which is transferred into convolutional neural network-long short-term memory (CNN-LSTM) to finally predict the future power demand. Experiments with the dataset of UCI household power consumption composed of a total of 2,075,259 time-series data show that the proposed model reduces the error by 23.63% and outperforms the state-of-the-art deep learning models including the CNN-LSTM. Especially, the proposed model improves the prediction performance by modeling the distribution of demand peaks in Euclidean space.

AB - In the time-series models for predicting residential energy consumption, the energy properties collected through multiple sensors usually include irregular and seasonal factors. The irregular pattern resulting from them is called peak demand, which is a major cause of performance degradation. In order to enhance the performance, we propose a convolutional-recurrent triplet network to learn and detect the demand peaks. The proposed model generates the latent space for demand peaks from data, which is transferred into convolutional neural network-long short-term memory (CNN-LSTM) to finally predict the future power demand. Experiments with the dataset of UCI household power consumption composed of a total of 2,075,259 time-series data show that the proposed model reduces the error by 23.63% and outperforms the state-of-the-art deep learning models including the CNN-LSTM. Especially, the proposed model improves the prediction performance by modeling the distribution of demand peaks in Euclidean space.

UR - http://www.scopus.com/inward/record.url?scp=85101360555&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85101360555&partnerID=8YFLogxK

U2 - 10.1109/ICDMW51313.2020.00110

DO - 10.1109/ICDMW51313.2020.00110

M3 - Conference contribution

AN - SCOPUS:85101360555

T3 - IEEE International Conference on Data Mining Workshops, ICDMW

SP - 757

EP - 761

BT - Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020

A2 - Di Fatta, Giuseppe

A2 - Sheng, Victor

A2 - Cuzzocrea, Alfredo

A2 - Zaniolo, Carlo

A2 - Wu, Xindong

PB - IEEE Computer Society

T2 - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020

Y2 - 17 November 2020 through 20 November 2020

ER -

Moon HJ, Bu SJ, Cho SB. Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network. In Di Fatta G, Sheng V, Cuzzocrea A, Zaniolo C, Wu X, editors, Proceedings - 20th IEEE International Conference on Data Mining Workshops, ICDMW 2020. IEEE Computer Society. 2020. p. 757-761. 9346355. (IEEE International Conference on Data Mining Workshops, ICDMW). doi: 10.1109/ICDMW51313.2020.00110

Learning Disentangled Representation of Residential Power Demand Peak via Convolutional-Recurrent Triplet Network

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this