Coordinating Principal Component Analyzers
Mixtures of Principal Component Analyzers can be used to model high dimensional data that lie on or near a low dimensional manifold. By linearly mapping the PCA subspaces to one global low dimensional space, we obtain a ‘global’ low dimensional coordinate system for the data. As shown by Roweis et al., ensuring consistent global low-dimensional coordinates for the data can be expressed as a penalized likelihood optimization problem. We show that a restricted form of the Mixtures of Probabilistic PCA model allows for a more efficient algorithm. Experimental results are provided to illustrate the viability method.
KeywordsDensity Model Informatics Institute Gaussian Density Feature Extraction Technique Probabilistic Principal Component Analyzer
Unable to display preview. Download preview PDF.
- 1.T.F. Cox and M.A.A. Cox. Multidimensional Scaling. Number 59 in Monographs on statistics and applied probability. Chapman & Hall, 1994.Google Scholar
- 2.Z. Ghahramani and G.E. Hinton. The EM Algorithm for Mixtures of Factor Analyzers. Technical Report CRG-TR-96-1, University of Toronto, Canada, 1996.Google Scholar
- 4.R.M. Neal and G.E. Hinton. A view of the EM algorithm that justifies incremental, sparse, and other variants. In M.I. Jorda, editor, Learning in Graphical Models, pages 355–368. Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998.Google Scholar
- 5.S.T. Roweis, L.K. Saul, and G.E. Hinton. Global coordination of local linear models. In T.G. Dietterich, S. Becker, and Z. Ghahramani, editors, Advances in Neural Information Processing Systems 14. MIT Press, 2002.Google Scholar
- 8.J.J. Verbeek, N. Vlassis, and B. Kröse. The Generative Self-Organizing Map: A Probabilistic Generalization of Kohonen’s SOM. Technical Report IAS-UVA-02-03, Informatics Institute, University of Amsterdam, The Netherlands, May 2002.Google Scholar
- 9.J.J. Verbeek, N. Vlassis, and B. Kröse. Procrustes Analysis to Coordinate Mixtures of Probabilistic Principal Component Analyzers. Technical report, Informatics Institute, University of Amsterdam, The Netherlands, February 2002.Google Scholar
- 10.N. Vlassis, Y. Motomura, and B. Kröse. Supervised dimension reduction of intrinsically low-dimensional data.Neural Computation, 14(1):191–215, January 2002.Google Scholar