Allied fuzzy c-means clustering using kernel methods

Allied fuzzy c-means (AFCM) clustering is a hybrid fuzzy clustering algorithm based on the combination of fuzzy c-means (FCM) and new possibilistic c-means (NPCM). AFCM can deal with noisy data better than FCM and does not generate coincident clusters. With kernel methods AFCM is improved as its kernel learning machine model. This proposed algorithm is called kernel allied fuzzy c-means (KAFCM) clustering. KAFCM is suitable for classification of nonlinear separable patterns while AFCM deals with linear separable patterns well. KAFCM can nonlinearly map the input data into a high-dimensional feature space where the nonlinear pattern now appears linear and AFCM is performed. The better performance of our proposed algorithm is shown by performing experiments on artificial dataset and standard IRIS dataset.