On linguistic-based clustering
Akira Sugawara, Naohiko Kinoshita, Yasunori Endo · 2014
Clustering which is one of data mining techniques is a method automatically classifying data into some clusters. Various types of clustering methods based on mathematical models are proposed. We call these methods model-based clustering. We use clustering methods to know data structure. However, we do not know which methods we should select unless we know data structure. Therefore, we propose a new clustering method based on linguistic rules, that is, fuzzy reasoning. We call the new method linguistic-based clustering. It is available when we do not even know data structure. Moreover, the effectiveness is shown through numerical examples.