Hybrid Possibilistic Clustering Algorithm

摘要

Possibilistic c-means (PCM) clustering can deal with noisy data better than fuzzy c-means (FCM) clustering. But PCM is very sensitive to good initialization and it has an undesirable tendency to produce coincident clusters. Improved possibilistic c-means (IPCM) algorithm tends to combine the benefits of PCM and FCM. However, the performance of IPCM depends heavily on the parameters. IPCM must compute the parameters twice, so it is a time-consuming algorithm. To deal with this problem, we adopt a technique called the possibilistic clustering algorithm (PCA). And then a hybrid possibilistic clustering algorithm (HPCA) is proposed to improve IPCM. Our experimental results show that the proposed algorithm compares favorably with FCM and IPCM.