首页> 外文会议>Proceedings of the International Colloquium on Information Fusion 2007 >Unified Description and Extension of Combination Rules in Evidence Reasoning Based on the Random Set Method
【24h】

Unified Description and Extension of Combination Rules in Evidence Reasoning Based on the Random Set Method

机译:基于随机集方法的证据推理中组合规则的统一描述和扩展

获取原文
获取外文期刊封面目录资料

摘要

The most popular rule of evidence combination, called Dempster's combination rule, has been widely used since it is a simple and convenient mechanism for combining multisource information. However,it has restrictive requirements and sometimes gives some counterintuitive results. With the desire to over-come its inherent limitations, many attempts have been made,alternative rules to the Dempster's combination rule can be found in the existing literature, but almost each of them can only be applied to some specific case. So it is very necessary to search their common foundation so as to create more new rules for combining available multisource information in more cases. In this paper, a unified formulation to describe most classical combination rules is presented based on the random set theory, and what's more, it provides us with new ideas for constructing other applicable combination rules.Finally, by use of this formulation, a new combination rule is constructed for overcoming some counterintuitive phenomena.
机译:最流行的证据组合规则称为Dempster组合规则,由于它是一种组合多源信息的简单便捷的机制,因此已被广泛使用。但是,它有严格的要求,有时会产生一些违反直觉的结果。为了克服其固有的局限性,已经进行了许多尝试,在现有文献中可以找到关于登普斯特组合规则的替代规则,但是几乎每个规则都只能应用于某些特定情况。因此,有必要寻找它们的共同基础,以便在更多情况下创建更多新规则以结合可用的多源信息。本文基于随机集理论提出了描述大多数经典组合规则的统一表述,并且为我们提供了构建其他适用组合规则的新思路。构造规则是为了克服一些违反直觉的现象。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号