Skip to main navigation Skip to search Skip to main content

An improved D-S algorithm under the new measure criteria of evidence conflict

  • Chang Hua Hu
  • , Xiao Sheng Si
  • , Zhi Jie Zhou
  • , Peng Wang

Research output: Contribution to journalArticlepeer-review

50 Scopus citations

Abstract

To overcome the shortcomings of conventional measure criterion of evidence conflict and solve the invalidation problem of Dempster-Shafer evidence combination rule with high conflict, a novel measure criterion of evidence conflict is defined through pignistic transformation. Based on this definition, a new improved D-S algorithm is proposed in this paper. According to the decision idea that the minority is subordinate to the majority, the proposed algorithm first preprocesses evidence through introducing weight coefficients which represent the importance degree of evidence, then combines preprocessed evidence using Dempster's rule. The numerical study shows that, compared with other methods, the new algorithm has made good performance in dealing with evidence conflict and can accelerate convergence and reduce the decision risk.

Original languageEnglish
Pages (from-to)1578-1583
Number of pages6
JournalTien Tzu Hsueh Pao/Acta Electronica Sinica
Volume37
Issue number7
StatePublished - Jul 2009
Externally publishedYes

Keywords

  • Combination rule
  • D-S theory
  • Evidence conflict
  • Information entropy
  • Pignistic transformation

Fingerprint

Dive into the research topics of 'An improved D-S algorithm under the new measure criteria of evidence conflict'. Together they form a unique fingerprint.

Cite this