05364597

a

05364597 Liu, Li Zhou, Jianzhong An, Xueli Li, Yinghai Liu, Qiang Improved fuzzy clustering method based on entropy coefficient and its application. Sun, Fuchun (ed.) et al., Advances in neural networks -- ISNN 2008. 5th international symposium on neural networks, ISNN 2008, Beijing, China, September 24--28, 2008. Proceedings, Part II. Berlin: Springer (ISBN 978-3-540-87733-2/pbk). Lecture Notes in Computer Science 5264, 11-20 (2008). 2008 Berlin: Springer EN fuzzy clustering entropy coefficient membership function weight

doi:10.1007/978-3-540-87734-9_2

Summary: Based on the principle of fuzzy clustering analysis and the theory of entropy, an improved fuzzy clustering method is given by improving the method of establishing the membership function, combining the clustering weight with the entropy coefficient, and replacing the Zadeh operator $M(\bigvee,\bigwedge$) with the weight average operator $M(\pm , \bullet )$. With the improved method, the zeroweight problem is addressed effectively, the weights of each factor are modified properly and the phenomenon of Major Factor Dominating is also alleviated appropriately. Finally, an illustrative example is given to clarify the method, which shows that the improved fuzzy clustering method is reasonable, feasible, simple and practical.