GTΔ

Detecting temporal changes in group stochastic processes

Edward Toth, Sanjay Chawla

Research output: Contribution to journalArticle

Abstract

Given a portfolio of stocks or a series of frames in a video how do we detect significant changes in a group of values for real-time applications? In this article, we formalize the problem of sequentially detecting temporal changes in a group of stochastic processes. As a solution to this particular problem, we propose the group temporal change (GTΔ) algorithm, a simple yet effective technique for the sequential detection of significant changes in a variety of statistical properties of a group over time. Due to the flexible framework of the GTΔ algorithm, a domain expert is able to select one or more statistical properties that they are interested in monitoring. The usefulness of our proposed algorithm is also demonstrated against state-of-the-art techniques on synthetically generated data as well as on two real-world applications; a portfolio of healthcare stocks over a 20 year period and a video monitoring the activity of our Sun.

Original languageEnglish
Article numbera39
JournalACM Transactions on Knowledge Discovery from Data
Volume12
Issue number4
DOIs
Publication statusPublished - 1 Jul 2018

Fingerprint

Random processes
Monitoring

Keywords

  • Anomaly detection
  • Group change detection
  • Time series analysis

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

GTΔ : Detecting temporal changes in group stochastic processes. / Toth, Edward; Chawla, Sanjay.

In: ACM Transactions on Knowledge Discovery from Data, Vol. 12, No. 4, a39, 01.07.2018.

Research output: Contribution to journalArticle

@article{7c480e540f264882bdf0d1ac9a0fe04b,
title = "GTΔ: Detecting temporal changes in group stochastic processes",
abstract = "Given a portfolio of stocks or a series of frames in a video how do we detect significant changes in a group of values for real-time applications? In this article, we formalize the problem of sequentially detecting temporal changes in a group of stochastic processes. As a solution to this particular problem, we propose the group temporal change (GTΔ) algorithm, a simple yet effective technique for the sequential detection of significant changes in a variety of statistical properties of a group over time. Due to the flexible framework of the GTΔ algorithm, a domain expert is able to select one or more statistical properties that they are interested in monitoring. The usefulness of our proposed algorithm is also demonstrated against state-of-the-art techniques on synthetically generated data as well as on two real-world applications; a portfolio of healthcare stocks over a 20 year period and a video monitoring the activity of our Sun.",
keywords = "Anomaly detection, Group change detection, Time series analysis",
author = "Edward Toth and Sanjay Chawla",
year = "2018",
month = "7",
day = "1",
doi = "10.1145/3183346",
language = "English",
volume = "12",
journal = "ACM Transactions on Knowledge Discovery from Data",
issn = "1556-4681",
publisher = "Association for Computing Machinery (ACM)",
number = "4",

}

TY - JOUR

T1 - GTΔ

T2 - Detecting temporal changes in group stochastic processes

AU - Toth, Edward

AU - Chawla, Sanjay

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Given a portfolio of stocks or a series of frames in a video how do we detect significant changes in a group of values for real-time applications? In this article, we formalize the problem of sequentially detecting temporal changes in a group of stochastic processes. As a solution to this particular problem, we propose the group temporal change (GTΔ) algorithm, a simple yet effective technique for the sequential detection of significant changes in a variety of statistical properties of a group over time. Due to the flexible framework of the GTΔ algorithm, a domain expert is able to select one or more statistical properties that they are interested in monitoring. The usefulness of our proposed algorithm is also demonstrated against state-of-the-art techniques on synthetically generated data as well as on two real-world applications; a portfolio of healthcare stocks over a 20 year period and a video monitoring the activity of our Sun.

AB - Given a portfolio of stocks or a series of frames in a video how do we detect significant changes in a group of values for real-time applications? In this article, we formalize the problem of sequentially detecting temporal changes in a group of stochastic processes. As a solution to this particular problem, we propose the group temporal change (GTΔ) algorithm, a simple yet effective technique for the sequential detection of significant changes in a variety of statistical properties of a group over time. Due to the flexible framework of the GTΔ algorithm, a domain expert is able to select one or more statistical properties that they are interested in monitoring. The usefulness of our proposed algorithm is also demonstrated against state-of-the-art techniques on synthetically generated data as well as on two real-world applications; a portfolio of healthcare stocks over a 20 year period and a video monitoring the activity of our Sun.

KW - Anomaly detection

KW - Group change detection

KW - Time series analysis

UR - http://www.scopus.com/inward/record.url?scp=85052576732&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052576732&partnerID=8YFLogxK

U2 - 10.1145/3183346

DO - 10.1145/3183346

M3 - Article

VL - 12

JO - ACM Transactions on Knowledge Discovery from Data

JF - ACM Transactions on Knowledge Discovery from Data

SN - 1556-4681

IS - 4

M1 - a39

ER -