Clustering over‐dispersed data with mixed feature types

• DOI: 10.1002/sam.11369
• Publication Date: 2018-01-10T17:32:57Z
• Authors: Lu Wang; Dongxiao Zhu; Ming Dong
• Keywords: Feature (linguistics); Single-linkage clustering

Despite many data clustering methods are available, most of them uncover compactness or connectivity as the intrinsic structure unlabeled data. Very few approaches explicitly consider cluster size distribution, especially over‐dispersed (high variance), which may represent yet another important aspect structural information In this paper, we propose a novel joint mixture model framework to estimate distribution together with (density). Our is sufficiently flexible and general capture wide range distributions from mixed feature types. Experiments on synthetic real‐world demonstrate superior performance our approach in recovering hidden