Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

Abstract

Recent developments in computing and IoT technology have enabled the daily generation of enormous amounts of time series data. These time series have to be analyzed to create value. A fundamental type of analysis is to find temporal correlations between given sets of time series. To provide a robust method for solving this problem, several properties are desirable. First, the method should have a strong theoretical foundation. Second, since temporal correlations can occur at different temporal scales, e.g., sub-second versus weekly, it is important that the method is capable of discovering multitemporal scale correlations. Finally, the method should be efficient and scalable. This paper presents an approach to search for synchronous correlations in big time series that displays all three properties: the proposed method (i) utilizes the metric of mutual information from information theory, providing a strong theoretical foundation, (ii) is able to discover correlations at multiple temporal scales, and (iii) works in an efficient, bottom-up fashion, making it scalable to large datasets. Our experiments verify that the proposed approach can identify various types of correlation relations across multiple temporal scales, while achieving a performance of an order of magnitude faster than the state-of-the-art techniques.

Original languageEnglish
Title of host publicationProceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019
Number of pages4
PublisherIEEE
Publication date8 Apr 2019
Pages1734-1737
Article number8731434
ISBN (Electronic)9781538674741
DOIs
Publication statusPublished - 8 Apr 2019
EventThe 35th IEEE International Conference on Data Engineering (ICDE) - Macau, Macau, China
Duration: 8 Apr 201912 Apr 2019

Conference

ConferenceThe 35th IEEE International Conference on Data Engineering (ICDE)
LocationMacau
CountryChina
CityMacau
Period08/04/201912/04/2019
SeriesProceedings of the International Conference on Data Engineering
ISSN1063-6382

Fingerprint

Time series
Information theory
Experiments

Keywords

  • Hill climbing
  • Mutual information
  • Sliding window
  • Temporal correlation

Cite this

Ho, T. T. N., Pedersen, T. B., Vu, M., Van, H. L., & Biscio, C. A. N. (2019). Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information. In Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019 (pp. 1734-1737). [8731434] IEEE. Proceedings of the International Conference on Data Engineering https://doi.org/10.1109/ICDE.2019.00185
Ho, Thi Thao Nguyen ; Pedersen, Torben Bach ; Vu, Mai ; Van, Ho Long ; Biscio, Christophe Ange Napoléon. / Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information. Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019. IEEE, 2019. pp. 1734-1737 (Proceedings of the International Conference on Data Engineering).
@inproceedings{607707c97e634ca3bb0c4d803715a7e4,
title = "Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information",
abstract = "Recent developments in computing and IoT technology have enabled the daily generation of enormous amounts of time series data. These time series have to be analyzed to create value. A fundamental type of analysis is to find temporal correlations between given sets of time series. To provide a robust method for solving this problem, several properties are desirable. First, the method should have a strong theoretical foundation. Second, since temporal correlations can occur at different temporal scales, e.g., sub-second versus weekly, it is important that the method is capable of discovering multitemporal scale correlations. Finally, the method should be efficient and scalable. This paper presents an approach to search for synchronous correlations in big time series that displays all three properties: the proposed method (i) utilizes the metric of mutual information from information theory, providing a strong theoretical foundation, (ii) is able to discover correlations at multiple temporal scales, and (iii) works in an efficient, bottom-up fashion, making it scalable to large datasets. Our experiments verify that the proposed approach can identify various types of correlation relations across multiple temporal scales, while achieving a performance of an order of magnitude faster than the state-of-the-art techniques.",
keywords = "Hill climbing, Mutual information, Sliding window, Temporal correlation",
author = "Ho, {Thi Thao Nguyen} and Pedersen, {Torben Bach} and Mai Vu and Van, {Ho Long} and Biscio, {Christophe Ange Napol{\'e}on}",
year = "2019",
month = "4",
day = "8",
doi = "10.1109/ICDE.2019.00185",
language = "English",
series = "Proceedings of the International Conference on Data Engineering",
publisher = "IEEE",
pages = "1734--1737",
booktitle = "Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019",
address = "United States",

}

Ho, TTN, Pedersen, TB, Vu, M, Van, HL & Biscio, CAN 2019, Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information. in Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019., 8731434, IEEE, Proceedings of the International Conference on Data Engineering, pp. 1734-1737, The 35th IEEE International Conference on Data Engineering (ICDE), Macau, China, 08/04/2019. https://doi.org/10.1109/ICDE.2019.00185

Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information. / Ho, Thi Thao Nguyen; Pedersen, Torben Bach; Vu, Mai; Van, Ho Long; Biscio, Christophe Ange Napoléon.

Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019. IEEE, 2019. p. 1734-1737 8731434 (Proceedings of the International Conference on Data Engineering).

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information

AU - Ho, Thi Thao Nguyen

AU - Pedersen, Torben Bach

AU - Vu, Mai

AU - Van, Ho Long

AU - Biscio, Christophe Ange Napoléon

PY - 2019/4/8

Y1 - 2019/4/8

N2 - Recent developments in computing and IoT technology have enabled the daily generation of enormous amounts of time series data. These time series have to be analyzed to create value. A fundamental type of analysis is to find temporal correlations between given sets of time series. To provide a robust method for solving this problem, several properties are desirable. First, the method should have a strong theoretical foundation. Second, since temporal correlations can occur at different temporal scales, e.g., sub-second versus weekly, it is important that the method is capable of discovering multitemporal scale correlations. Finally, the method should be efficient and scalable. This paper presents an approach to search for synchronous correlations in big time series that displays all three properties: the proposed method (i) utilizes the metric of mutual information from information theory, providing a strong theoretical foundation, (ii) is able to discover correlations at multiple temporal scales, and (iii) works in an efficient, bottom-up fashion, making it scalable to large datasets. Our experiments verify that the proposed approach can identify various types of correlation relations across multiple temporal scales, while achieving a performance of an order of magnitude faster than the state-of-the-art techniques.

AB - Recent developments in computing and IoT technology have enabled the daily generation of enormous amounts of time series data. These time series have to be analyzed to create value. A fundamental type of analysis is to find temporal correlations between given sets of time series. To provide a robust method for solving this problem, several properties are desirable. First, the method should have a strong theoretical foundation. Second, since temporal correlations can occur at different temporal scales, e.g., sub-second versus weekly, it is important that the method is capable of discovering multitemporal scale correlations. Finally, the method should be efficient and scalable. This paper presents an approach to search for synchronous correlations in big time series that displays all three properties: the proposed method (i) utilizes the metric of mutual information from information theory, providing a strong theoretical foundation, (ii) is able to discover correlations at multiple temporal scales, and (iii) works in an efficient, bottom-up fashion, making it scalable to large datasets. Our experiments verify that the proposed approach can identify various types of correlation relations across multiple temporal scales, while achieving a performance of an order of magnitude faster than the state-of-the-art techniques.

KW - Hill climbing

KW - Mutual information

KW - Sliding window

KW - Temporal correlation

UR - http://www.scopus.com/inward/record.url?scp=85068016582&partnerID=8YFLogxK

U2 - 10.1109/ICDE.2019.00185

DO - 10.1109/ICDE.2019.00185

M3 - Article in proceeding

T3 - Proceedings of the International Conference on Data Engineering

SP - 1734

EP - 1737

BT - Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019

PB - IEEE

ER -

Ho TTN, Pedersen TB, Vu M, Van HL, Biscio CAN. Efficient Bottom-Up Discovery of Multi-Scale Time Series Correlations Using Mutual Information. In Proceedings - 2019 IEEE 35th International Conference on Data Engineering, ICDE 2019. IEEE. 2019. p. 1734-1737. 8731434. (Proceedings of the International Conference on Data Engineering). https://doi.org/10.1109/ICDE.2019.00185