On Fuzzy Clustering for Incomplete Spherical Data and for Incomplete Multivariate Categorical Data

摘要

In this paper, six fuzzy clustering algorithms for incomplete data are proposed that use the optimal completion strategy, three of which are for incomplete spherical data and these of which are for incomplete categorical multivariate data. In numerical experiments using a real dataset, each of the proposed methods outperformed its counterpart method based on the whole data strategy (WDS) in terms of accuracy.