计算机科学 ›› 2009, Vol. 36 ›› Issue (8): 260-263.

• 人工智能 • 上一篇    下一篇

一种DSm理论相关证据模型

王进,孙怀江   

  1. (南京理工大学计算机系 南京 210094)
  • 出版日期:2018-11-16 发布日期:2018-11-16

Model for Dependent Evidences in DSmT Framework

WANG Jin,SUN Huai-Jiang   

  • Online:2018-11-16 Published:2018-11-16

摘要: DSm理论(Dezert Smarandache Theory,DSmT)作为D-S (Dempster-Shaf er)证据理论的推广,在进行多个证据合成时同样要求证据相互独立,这是一个非常苛刻的条件,在实际中很难满足。提出了在DSm理论框架中表示相关证据的一种模型,其中两个相关证据各由一个独立源证据和一个相关源证据正交和合成,相关证据的合成就归结为这两个独立源证据和这个相关源证据的正交和合成。为此就要先辨识出独立源证据。证明了在自由DSm模型中辫识结果的唯一条件,并针对条件不满足的情况提出了一种近似的

关键词: Dezert-Smarandache理论,不确定性,相关性

Abstract: DSmT(Dezert Smarandache Theory) which is an extension of Dempster-Shafer theory assumes that sources of evidences are independent as the same as Dempster-Shafer Evidence Theory.This is a very strict condition that is difficult to satisfy in practice. A new model used to represent interrelated evidences in DSmT framework was proposed. In this model, two dependent evidences were considered resulted from orthogonal sum of one dependent original evidence and two independent original evidences, respectively. Combining the two dependent evidences can be reduced to an orthogonal sum of the two independent original evidences and the dependent original evidence. For this, independent original evidences must be identified firstly. A condition was proven to guarantee that the identified dependent original evidence is unique in free DSm’model,and if this condition is not satisficd,an approximate solution is proposed.

Key words: Dezert-smarandache theory,Uncertainty,Dependence

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!