Title :
A kernel-based fuzzy c-means algorithm with partition index maximization
Author :
Tsai, Hsu-Shen ; Yang, Miin-Shen
Author_Institution :
Dept. of Manage. Inf. Syst., Takming Univ. of Sci. & Technol., Taipei, Taiwan
Abstract :
This paper presents a kernel-based fuzzy c-means algorithm with partition index maximization, called KPIM algorithm. The proposed KPIM algorithm is more robust than the partition index maximization algorithm proposed by Özdemir and Akarum. Experiments show that the advantage of KPIM are robust properties: (1) robust to fuzziness parameter m, (2) robust to outlier, (3) robust to image artifacts; and fast computational performance. Especially, KPIM can overcome drawbacks of PIM, and are well used in image segmentation.
Keywords :
fuzzy set theory; image segmentation; optimisation; pattern clustering; KPIM algorithm; image artifacts; image segmentation; kernel-based fuzzy c-means algorithm; partition index maximization algorithm; Clustering algorithms; Gaussian noise; Image segmentation; Indexes; Kernel; Partitioning algorithms; Robustness; Fuzzy c-means(FCM); Image segmentation; Kernel; Outlier; Partition index maximization;
Conference_Titel :
Fuzzy Systems and Knowledge Discovery (FSKD), 2010 Seventh International Conference on
Conference_Location :
Yantai, Shandong
Print_ISBN :
978-1-4244-5931-5
DOI :
10.1109/FSKD.2010.5569636
