Burst detection from multiple data streams

A network-based approach

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Modeling and detecting bursts in data streams is an important area of research with a wide range of applications. In this paper, we present a novel method to analyze and identify correlated burst patterns by considering multiple data streams that coevolve over time. The main technical contribution of our research is the use of a dynamic probabilistic network to model the dependency structures observed within these data streams. Such dependencies provide meaningful information concerning the overall system dynamics and should be explicitly integrated into the burst detection process. Using both synthetic scenarios and two real-world datasets, we compare our method with an existing burst-detection algorithm. Initial experimental results indicate that our approach allows for more balanced and accurate burst quantification.

Original languageEnglish (US)
Article number5378562
Pages (from-to)258-267
Number of pages10
JournalIEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews
Volume40
Issue number3
DOIs
StatePublished - May 2010

Fingerprint

Dynamical systems

Keywords

  • Burst detection
  • Factorial hidden Markov model (HMMs)
  • Multiple data streams
  • Probabilistic network

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Computer Science Applications
  • Human-Computer Interaction
  • Information Systems
  • Software

Cite this

@article{f742dc86e16f4971bd601b5428354ea9,
title = "Burst detection from multiple data streams: A network-based approach",
abstract = "Modeling and detecting bursts in data streams is an important area of research with a wide range of applications. In this paper, we present a novel method to analyze and identify correlated burst patterns by considering multiple data streams that coevolve over time. The main technical contribution of our research is the use of a dynamic probabilistic network to model the dependency structures observed within these data streams. Such dependencies provide meaningful information concerning the overall system dynamics and should be explicitly integrated into the burst detection process. Using both synthetic scenarios and two real-world datasets, we compare our method with an existing burst-detection algorithm. Initial experimental results indicate that our approach allows for more balanced and accurate burst quantification.",
keywords = "Burst detection, Factorial hidden Markov model (HMMs), Multiple data streams, Probabilistic network",
author = "Aaron Sun and Dajun Zeng and Hsinchun Chen",
year = "2010",
month = "5",
doi = "10.1109/TSMCC.2009.2037311",
language = "English (US)",
volume = "40",
pages = "258--267",
journal = "IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews",
issn = "1094-6977",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Burst detection from multiple data streams

T2 - A network-based approach

AU - Sun, Aaron

AU - Zeng, Dajun

AU - Chen, Hsinchun

PY - 2010/5

Y1 - 2010/5

N2 - Modeling and detecting bursts in data streams is an important area of research with a wide range of applications. In this paper, we present a novel method to analyze and identify correlated burst patterns by considering multiple data streams that coevolve over time. The main technical contribution of our research is the use of a dynamic probabilistic network to model the dependency structures observed within these data streams. Such dependencies provide meaningful information concerning the overall system dynamics and should be explicitly integrated into the burst detection process. Using both synthetic scenarios and two real-world datasets, we compare our method with an existing burst-detection algorithm. Initial experimental results indicate that our approach allows for more balanced and accurate burst quantification.

AB - Modeling and detecting bursts in data streams is an important area of research with a wide range of applications. In this paper, we present a novel method to analyze and identify correlated burst patterns by considering multiple data streams that coevolve over time. The main technical contribution of our research is the use of a dynamic probabilistic network to model the dependency structures observed within these data streams. Such dependencies provide meaningful information concerning the overall system dynamics and should be explicitly integrated into the burst detection process. Using both synthetic scenarios and two real-world datasets, we compare our method with an existing burst-detection algorithm. Initial experimental results indicate that our approach allows for more balanced and accurate burst quantification.

KW - Burst detection

KW - Factorial hidden Markov model (HMMs)

KW - Multiple data streams

KW - Probabilistic network

UR - http://www.scopus.com/inward/record.url?scp=77951171882&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951171882&partnerID=8YFLogxK

U2 - 10.1109/TSMCC.2009.2037311

DO - 10.1109/TSMCC.2009.2037311

M3 - Article

VL - 40

SP - 258

EP - 267

JO - IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews

JF - IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews

SN - 1094-6977

IS - 3

M1 - 5378562

ER -