摘要
针对决策信息信息表中新对象的分类问题,提出一种基于改进的贝叶斯粗糙集和证据理论的决策信息融合方法。对传统的贝叶斯粗糙集进行改进,扩展到多决策类,定义了支持度的概念以此反映确切分类的对象所占的百分比。利用贝叶斯粗糙集的支持度和置信增益函数作为证据的支持程度,得到各准则下的证据基本概率分配函数,并利用证据合成法则对多个证据进行合成,以此进行决策。将上述方法应用于设备故障的诊断问题中,通过方法的对比验证了该方法实践应用的有效性。
According to the classification problem of the new object in the decision information table, this paper constructed a method of decision information fusion based on the theory of Bayesian rough set and evidence theory. To deal with the problem of multiple decision classes, it improved the traditional Bayesian rough set theory. It defined the support degree to express the exact classification. Then it calculated the approximation classified quality and the certainty gain function of the Bayesian rough set to rate the support degree of the evidences, and used the normalization method to construct the basic probability as- signments. It fused the evidences by using the D-S combination rule. Finally, it applied the proposed method to the problems of equipment fault diagnosis, and the results show the effectiveness of practical application of this method.
出处
《计算机应用研究》
CSCD
北大核心
2014年第9期2625-2628,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(71271216)
国家社科基金重点项目(12AZD109)
关键词
贝叶斯粗糙集
证据理论
证据权重
多准则决策
Bayesian rough set
evidence theory
evidence weight
multi-criteria decision-making
