A Novel Approach to Feature Selection for Clustering

Feature selection has received considerable attentions in various areas as a way to select informative features and to simplify the statistical model through dimensional reduction. In this work, we introduce a novel concept, membership probability of a feature, and propose a novel approach to feature selection for clustering which can find the most optimal candidate features effectively among the original feature space. The efficiency and effectiveness of our approach is demonstrated through extensive comparisons with other methods using real-world data of high dimensionality.