PCA-guided fuzzy cluster validation with noise rejection

摘要: This paper considers cluster validation for fuzzy clustering with noise rejection. Although rejection mechanisms such as or graded possibilistic make it possible to remove the influence of noisy samples, they also create problems in applying conventional validity measures designed probabilistic constraints. In this paper, a PCA-guided approach is developed, which rotated optimal indicator derived manner, considering responsibility weights c-means clustering. The deviation between current solution and estimated through procrustean transformation. Several experimental results demonstrate that proposed works well selecting both initialization number.