Fuzzy clustering for uncertainty data
Mika Sato‐Ilic · 2003
This paper proposes a clustering model which can capture the change of vagueness included in data when the data is observed through several times and the vagueness is changed according to the times. In this paper, the vagueness is treated as fuzzy data, that is, it is defined as convex normal fuzzy sets. Due to the definitions of the different vagueness of each observation, the dissimilarity (or similarity) between a pair of objects has the property of asymmetric relation. This numerical example shows the validity of the model.