CSDL Home IEEE Transactions on Pattern Analysis & Machine Intelligence 2009 vol.31 Issue No.05 - May
Issue No.05 - May (2009 vol.31)
Yuanhong Li , Wayne State University, Detroit
Ming Dong , Wayne State University, Detroit
Jing Hua , Wayne State University, Detroit
In this paper, we propose a novel approach of simultaneous localized feature selection and model detection for unsupervised learning. In our approach, local feature saliency, together with other parameters of Gaussian mixtures, are estimated by Bayesian variational learning. Experiments performed on both synthetic and real-world data sets demonstrate that our approach is superior over both global feature selection and subspace clustering methods.
Unsupervised, localized, feature selection, Bayesian.
Yuanhong Li, Ming Dong, Jing Hua, "Simultaneous Localized Feature Selection and Model Detection for Gaussian Mixtures", IEEE Transactions on Pattern Analysis & Machine Intelligence, vol.31, no. 5, pp. 953-960, May 2009, doi:10.1109/TPAMI.2008.261