Feature Selection for Non Gaussian Mixture Models

We present in this paper a new approach for unsupervised feature selection for non Gaussian data controlled by a finite mixture of generalized Dirichlet distributions. We model each feature by a mixture of two Beta distributions: one relevant and depends on component labels while the second is uninformative for the clustering. The relevance of each feature is then quantified by the mixture weight associated to the relevant Beta distribution. Experiments in summarizing image collections have shown the effectiveness of our approach.