A Bayesian sampling approach to decision fusion using hierarchical models

Research output: Contribution to journalArticle

53 Citations (Scopus)

Abstract

Data fusion and distributed detection have been studied extensively, and numerous results have been obtained during the past two decades. In this paper, the design of fusion rule for distributed detection problems is re-examined, and a novel approach using Bayesian inference tools is proposed. Specifically, the decision fusion problem is reformulated using hierarchical models, and a Gibbs sampler is proposed to perform posterior probability-based fusion. Performancewise, it is essentially identical to the optimal likelihood-based fusion rule whenever it exists. The true merit of this approach is its applicability to various complex situations, e.g., in dealing with unknown signal/noise statistics where the likelihood-based fusion rule may not be easy to obtain or may not even exist.

Original languageEnglish (US)
Pages (from-to)1809-1818
Number of pages10
JournalIEEE Transactions on Signal Processing
Volume50
Issue number8
DOIs
StatePublished - Aug 2002

Fingerprint

Data fusion
Statistics
Sampling

Keywords

  • Bayesian inference
  • Decision fusion
  • Gibbs sampler
  • Hierarchical models

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing

Cite this

A Bayesian sampling approach to decision fusion using hierarchical models. / Chen, Biao; Varshney, Pramod Kumar.

In: IEEE Transactions on Signal Processing, Vol. 50, No. 8, 08.2002, p. 1809-1818.

Research output: Contribution to journalArticle

@article{11422c92d3074f2e85d2d5f72a41b371,
title = "A Bayesian sampling approach to decision fusion using hierarchical models",
abstract = "Data fusion and distributed detection have been studied extensively, and numerous results have been obtained during the past two decades. In this paper, the design of fusion rule for distributed detection problems is re-examined, and a novel approach using Bayesian inference tools is proposed. Specifically, the decision fusion problem is reformulated using hierarchical models, and a Gibbs sampler is proposed to perform posterior probability-based fusion. Performancewise, it is essentially identical to the optimal likelihood-based fusion rule whenever it exists. The true merit of this approach is its applicability to various complex situations, e.g., in dealing with unknown signal/noise statistics where the likelihood-based fusion rule may not be easy to obtain or may not even exist.",
keywords = "Bayesian inference, Decision fusion, Gibbs sampler, Hierarchical models",
author = "Biao Chen and Varshney, {Pramod Kumar}",
year = "2002",
month = "8",
doi = "10.1109/TSP.2002.800419",
language = "English (US)",
volume = "50",
pages = "1809--1818",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "8",

}

TY - JOUR

T1 - A Bayesian sampling approach to decision fusion using hierarchical models

AU - Chen, Biao

AU - Varshney, Pramod Kumar

PY - 2002/8

Y1 - 2002/8

N2 - Data fusion and distributed detection have been studied extensively, and numerous results have been obtained during the past two decades. In this paper, the design of fusion rule for distributed detection problems is re-examined, and a novel approach using Bayesian inference tools is proposed. Specifically, the decision fusion problem is reformulated using hierarchical models, and a Gibbs sampler is proposed to perform posterior probability-based fusion. Performancewise, it is essentially identical to the optimal likelihood-based fusion rule whenever it exists. The true merit of this approach is its applicability to various complex situations, e.g., in dealing with unknown signal/noise statistics where the likelihood-based fusion rule may not be easy to obtain or may not even exist.

AB - Data fusion and distributed detection have been studied extensively, and numerous results have been obtained during the past two decades. In this paper, the design of fusion rule for distributed detection problems is re-examined, and a novel approach using Bayesian inference tools is proposed. Specifically, the decision fusion problem is reformulated using hierarchical models, and a Gibbs sampler is proposed to perform posterior probability-based fusion. Performancewise, it is essentially identical to the optimal likelihood-based fusion rule whenever it exists. The true merit of this approach is its applicability to various complex situations, e.g., in dealing with unknown signal/noise statistics where the likelihood-based fusion rule may not be easy to obtain or may not even exist.

KW - Bayesian inference

KW - Decision fusion

KW - Gibbs sampler

KW - Hierarchical models

UR - http://www.scopus.com/inward/record.url?scp=0036683910&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036683910&partnerID=8YFLogxK

U2 - 10.1109/TSP.2002.800419

DO - 10.1109/TSP.2002.800419

M3 - Article

AN - SCOPUS:0036683910

VL - 50

SP - 1809

EP - 1818

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 8

ER -