基于诱导覆盖的粗糙直觉模糊集模型

A new type of covering based generalized rough intuitionistic fuzzy set model

基于粗糙集和直觉模糊集的信息融合方法是数据挖掘领域的一个热点研究课题.目前众多的融合方法主要针对的是单一覆盖背景下的数据,利用覆盖知识中对元素的定性描述来构造覆盖粗糙直觉模糊集模型.但具有多属性的数据形成的是多个覆盖构成的覆盖族,而已有的覆盖粗糙直觉模糊集模型在处理这类数据具有一定的局限性.在多种属性形成的覆盖族上,利用诱导覆盖的概念,从元素间的相似度的角度出发,建立了覆盖族上基于诱导覆盖的粗糙直觉模糊集模型.讨论了该模型的一些重要性质,同时提出了模糊粗糙度概念和相似度的概念,并在此讨论了新模型的不确定性度量.最后,利用算例说明了新模型及其度量指标在现实具体问题中的有效应用.

Information fusion of rough sets and intuitionistic fuzzy sets is a hot research topic in the field of data mining.At present,existing fusion methods are oriented to data with a single covering and employ the qualitative description of objects in a covering knowledge to construct the models of the covering rough intuitionistic fuzzy set.However,data with multiple attributes generate covering family with multiple coverings,while existing the covering rough intuitionistic fuzzy set models have some limitations in dealing with those data.In consideration of data with multiple attributes,this paper uses the concept of covering induction to introduce the covering rough intuitionistic fuzzy set models in covering family from the view point of similarity of objects.Some important properties of the present model are further discussed.Meanwhile,the notions of fuzzy roughness and similarity are proposed,based on which the uncertainty measures of the new models are investigated.Finally,the efficiency of the new models and the new measures are examined by using practical examples in real life.