Minor Probability Events' Detection in Big Data: An Integrated Approach with Bayes Detection and MIM

Shuo Wan, Jiaxun Lu, Pingyi Fan, Khaled Letaief

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

In some common scenarios, rare events tend to be more important and worth careful detection. In this case, minor probability event detection rises up as an important issue in big data analytics. With the existence of empirical prior probability distributions of events, one common method to minimize the average detection error is Bayes detection. However, in extreme cases where the prior probability is small, the Bayes method may result in high miss detection rate. To overcome this problem, we proposed a modified detection algorithm by integrating the Bayes method with the message importance measure. Simulation results show that the algorithm significantly reduces the miss detection rate of extremely rare events, which may be helpful to dig rarely occurring phenomenons in big data.

Original languageEnglish
Article number8629291
Pages (from-to)418-421
Number of pages4
JournalIEEE Communications Letters
Volume23
Issue number3
DOIs
Publication statusPublished - 1 Mar 2019
Externally publishedYes

    Fingerprint

Keywords

  • false alarm rate
  • Message importance
  • minor probability
  • miss detection rate

ASJC Scopus subject areas

  • Modelling and Simulation
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this