关键词: 核, 模糊聚类, 模式识别, 可能性聚类

Abstract: A possibilistic clustering algorithm reduces the noisy influence on cluster centers by producing possibilities.However it often tends to find the identical cluster.To overcome this shortcoming，PFCM is proposed，which divides the data set into [k] different clusters through producing memberships and possibilities simultaneously，along with the cluster centers.But when two highly unequal sized clusters are given，PFCM fails to give the desired results.Therefore，a mercer-kernel based mixed C-means fuzzy clustering algorithm WKFM with attributes weights in feature space is proposed，which considers the imbalance between the attributes fully and uses kernel function to make it possible to cluster data that is linearly non-separable in the original space into homogeneous groups in the transformed high dimensional space with optimized kernel parameters.The experimental results show that the proposed algorithm can precisely find the ideal cluster centers and gives better results.

Key words: kernel, fuzzy clustering, pattern recognition, possibilistic clustering