Information granulation: a search for data structures

Pedrycz, W. and Bargiela, A., 2001. Information granulation: a search for data structures. In: KES 2001. 5th International Conference on Knowledge Based Intelligent Information Engineering Systems and Allied Technology, Osaka, Japan, 6-8 September 2001, Osaka, Japan.

Full text not available from this repository.

Abstract

Revealing a structure in data is of paramount importance in a broad range of problems of information processing. In spite of the specificity of the problem in which such analysis is realized, there is an evident commonality of all these pursuits worth emphasizing. One can distinguish between a core and a residual of the data structure. In this study, we propose a formal environment supporting these concepts and develop its algorithmic fabric. The algorithms leading to the development of information granules lend themselves to the Fuzzy C-Means (FCM) equipped with the Tchebyschev (1_∞) metric. The paper offers a novel contribution of a gradient-based learning of the prototypes developed in this form of clustering. The 1_∞ metric promotes a design of easily interpretable hyperboxes. In this setting, we quantify the notion of the core and residual part of the data. An interaction between information granules is discussed as well

Item Type:	Conference contribution
Creators:	Pedrycz, W. and Bargiela, A.
Publisher:	IOS Press
Place of Publication:	Amsterdam, Netherlands
Date:	2001
Volume:	2
Divisions:	Schools > School of Science and Technology
Record created by:	EPrints Services
Date Added:	09 Oct 2015 09:45
Last Modified:	19 Oct 2015 14:22
URI:	https://irep.ntu.ac.uk/id/eprint/2363

Actions (login required)

Edit View

Views

Downloads