A Novel Kernel Possibitistic Fuzzy C-Means Clustering Algorithm for Large Scale Data Sets
Kernel Method(KM) is a algorithm that, by replacing the inner product with an appropriate positive definite function, implicitly perform a nonlinear mapping of the input data into a high-dimensional feature space. The incorporation of KM enables the Kernel Possibitistic Fuzzy c-Means (KPFCM) algorithm to explore the inherent data pattern in the new space. However, the applications of KPFCM algorithm are confined to small scale data sets due to its expensive computation and storage cost. In this paper, KPFCM-L algorithm is presented to solve the large scale clustering problem. In KPFCM-L, kernel method is adopted to solve the nonlinear separable problem and get nonlinear boundaries. The proposed algorithm is applied to a customer segmentation application and the simulation results indicate the algorithm is very efficient for large scale data sets.
kernel method fuzzy c-means possibilistic fuzzy c-means large scale data sets
Yu Qu Hongye Su Ying Zhang Jian Chu
State Key Lab.Of Industrial Control Technology, Institute of Advanced Process Control, Zhejiang University
国际会议
武汉
英文
2007-09-21（万方平台首次上网日期，不代表论文的发表时间）