Representative clustering of uncertain data

摘要: This paper targets the problem of computing meaningful clusterings from uncertain data sets. Existing methods for clustering compute a single without any indication its quality and reliability; thus, decisions based on their results are questionable. In this paper, we describe framework, possible-worlds semantics; when applied an dataset, it computes set representative clusterings, each which has probabilistic guarantee not to exceed some maximum distance ground truth clustering, i.e., actual (but unknown) data. Our framework can be combined with existing algorithm is first provide guarantees about result. addition, our experimental evaluation shows that have much smaller deviation than approaches, thus reducing effect uncertainty.