Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

19 Downloads (Pure)

Resumé

Very large time series are increasingly available from an ever wider range of IoT-enabled sensors deployed in different environments. Significant insights and values can be obtained from these time series through performing cross-domain analyses, one of which is analyzing time delay temporal correlations across different datasets. Most existing works in this area are either limited in the type of detected relations, e.g., linear relations alone, only working with a fixed temporal scale, or not considering time delay between time series. This paper presents our Time delaY COrrelation Search (TYCOS) approach which provides a powerful and robust solution with the following features: (1) TYCOS is based on the concept of mutual information (MI) from information theory, giving it a strong theoretical foundation to detect all types of relations including non-linear ones, (2) TYCOS is able to discover time delay correlations at multiple temporal scales, (3) TYCOS works in an efficient, bottom-up fashion, pruning non-interesting time intervals from the search by employing a novel MI-based theory, and (4) TYCOS is designed to efficiently minimize computational redundancy. A comprehensive experimental evaluation using synthetic and real-world datasets from the energy and smart city domains shows that TYCOS is able to find significant time delay correlations across different time intervals among big time series. The performance evaluation shows that TYCOS can scale to large datasets, and achieve an average speedup of 2 to 3 orders of magnitude compared to the baselines by using the proposed optimizations.
OriginalsprogEngelsk
TitelProceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020
ForlagAssociation for Computing Machinery
ISBN (Elektronisk)978-3-89318-083-7
StatusAccepteret/In press - 2020

Fingerprint

Time series
Time delay
Information theory
Redundancy
Sensors

Citer dette

Ho, N. T. T., Pedersen, T. B., Ho, L. V., & Vu, M. (Accepteret/In press). Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series. I Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020 Association for Computing Machinery.
Ho, Nguyen Thi Thao ; Pedersen, Torben Bach ; Ho, Long Van ; Vu, Mai. / Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series. Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020. Association for Computing Machinery, 2020.
@inproceedings{35c5db200d904fb794d919a32a8a5c7a,
title = "Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series",
abstract = "Very large time series are increasingly available from an ever wider range of IoT-enabled sensors deployed in different environments. Significant insights and values can be obtained from these time series through performing cross-domain analyses, one of which is analyzing time delay temporal correlations across different datasets. Most existing works in this area are either limited in the type of detected relations, e.g., linear relations alone, only working with a fixed temporal scale, or not considering time delay between time series. This paper presents our Time delaY COrrelation Search (TYCOS) approach which provides a powerful and robust solution with the following features: (1) TYCOS is based on the concept of mutual information (MI) from information theory, giving it a strong theoretical foundation to detect all types of relations including non-linear ones, (2) TYCOS is able to discover time delay correlations at multiple temporal scales, (3) TYCOS works in an efficient, bottom-up fashion, pruning non-interesting time intervals from the search by employing a novel MI-based theory, and (4) TYCOS is designed to efficiently minimize computational redundancy. A comprehensive experimental evaluation using synthetic and real-world datasets from the energy and smart city domains shows that TYCOS is able to find significant time delay correlations across different time intervals among big time series. The performance evaluation shows that TYCOS can scale to large datasets, and achieve an average speedup of 2 to 3 orders of magnitude compared to the baselines by using the proposed optimizations.",
keywords = "time delay, temporal correlation, mutual information, hill climbing",
author = "Ho, {Nguyen Thi Thao} and Pedersen, {Torben Bach} and Ho, {Long Van} and Mai Vu",
year = "2020",
language = "English",
booktitle = "Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020",
publisher = "Association for Computing Machinery",
address = "United States",

}

Ho, NTT, Pedersen, TB, Ho, LV & Vu, M 2020, Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series. i Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020. Association for Computing Machinery.

Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series. / Ho, Nguyen Thi Thao; Pedersen, Torben Bach; Ho, Long Van; Vu, Mai.

Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020. Association for Computing Machinery, 2020.

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series

AU - Ho, Nguyen Thi Thao

AU - Pedersen, Torben Bach

AU - Ho, Long Van

AU - Vu, Mai

PY - 2020

Y1 - 2020

N2 - Very large time series are increasingly available from an ever wider range of IoT-enabled sensors deployed in different environments. Significant insights and values can be obtained from these time series through performing cross-domain analyses, one of which is analyzing time delay temporal correlations across different datasets. Most existing works in this area are either limited in the type of detected relations, e.g., linear relations alone, only working with a fixed temporal scale, or not considering time delay between time series. This paper presents our Time delaY COrrelation Search (TYCOS) approach which provides a powerful and robust solution with the following features: (1) TYCOS is based on the concept of mutual information (MI) from information theory, giving it a strong theoretical foundation to detect all types of relations including non-linear ones, (2) TYCOS is able to discover time delay correlations at multiple temporal scales, (3) TYCOS works in an efficient, bottom-up fashion, pruning non-interesting time intervals from the search by employing a novel MI-based theory, and (4) TYCOS is designed to efficiently minimize computational redundancy. A comprehensive experimental evaluation using synthetic and real-world datasets from the energy and smart city domains shows that TYCOS is able to find significant time delay correlations across different time intervals among big time series. The performance evaluation shows that TYCOS can scale to large datasets, and achieve an average speedup of 2 to 3 orders of magnitude compared to the baselines by using the proposed optimizations.

AB - Very large time series are increasingly available from an ever wider range of IoT-enabled sensors deployed in different environments. Significant insights and values can be obtained from these time series through performing cross-domain analyses, one of which is analyzing time delay temporal correlations across different datasets. Most existing works in this area are either limited in the type of detected relations, e.g., linear relations alone, only working with a fixed temporal scale, or not considering time delay between time series. This paper presents our Time delaY COrrelation Search (TYCOS) approach which provides a powerful and robust solution with the following features: (1) TYCOS is based on the concept of mutual information (MI) from information theory, giving it a strong theoretical foundation to detect all types of relations including non-linear ones, (2) TYCOS is able to discover time delay correlations at multiple temporal scales, (3) TYCOS works in an efficient, bottom-up fashion, pruning non-interesting time intervals from the search by employing a novel MI-based theory, and (4) TYCOS is designed to efficiently minimize computational redundancy. A comprehensive experimental evaluation using synthetic and real-world datasets from the energy and smart city domains shows that TYCOS is able to find significant time delay correlations across different time intervals among big time series. The performance evaluation shows that TYCOS can scale to large datasets, and achieve an average speedup of 2 to 3 orders of magnitude compared to the baselines by using the proposed optimizations.

KW - time delay

KW - temporal correlation

KW - mutual information

KW - hill climbing

M3 - Article in proceeding

BT - Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020

PB - Association for Computing Machinery

ER -

Ho NTT, Pedersen TB, Ho LV, Vu M. Efficient Search for Multi-Scale Time Delay Correlations in Big Time Series. I Proceedings - The 23rd International Conference on Extending Database Technology (EDBT), March 30-April 2, 2020. Association for Computing Machinery. 2020