Summary on Three-way Decision Method for Clustering Uncertainty
Due to the characteristic of complexity,diversity and uncertainty of data,clustering usually has uncertainty,which refers to the inconsistency of clustering results,thus affecting the accuracy of analysis.The Three-way Decision method can bring a series of advantages by constructing an uncertain domain and making delayed decision on the uncertain domain.In recent years,a lot of important work has emerged from this research hotspot.This paper introduces the model and the research direction of Three-way Decision method with clustering uncertainty,and summarizes the problems encountered in the construction process of Three-way Clustering.Firstly,the expression form of the Three-way Clustering is given.Secondly,the paper describes how to construct the core domain and boundary domain of the Three-way Clustering,how to determine the threshold value and how to design the validity index,respectively.It also introduces the representative work among them.Finally,the paper summarizes the Three-way Decision method of clustering uncertainty,and looks forward to the future research direction.
clusteringuncertaintyThree-way Clusteringcore domainboundary domainthreshold valuevalidity index