Auxiliary Variational Information Maximization for Dimensionality Reduction
Mutual Information (MI) is a long studied measure of information content, and many attempts to apply it to feature extraction and stochastic coding have been made. However, in general MI is computationally intractable to evaluate, and most previous studies redefine the criterion in forms of approximations. Recently we described properties of a simple lower bound on MI, and discussed its links to some of the popular dimensionality reduction techniques . Here we introduce a richer family of auxiliary variational bounds on MI, which generalizes our previous approximations. Our specific focus then is on applying the bound to extracting informative lower-dimensional projections in the presence of irreducible Gaussian noise. We show that our method produces significantly tighter bounds than the well-known as-if Gaussian approximations of MI. We also show that the auxiliary variable method may help to significantly improve on reconstructions from noisy lower-dimensional projections.
KeywordsMutual Information Dimensionality Reduction Auxiliary Variable Conditional Entropy Rich Family
Unable to display preview. Download preview PDF.
- 1.Barber, D., Agakov, F.V.: The IM Algorithm: A Variational Approach to Information Maximization. In: NIPS, MIT Press, Cambridge (2003)Google Scholar
- 2.Linsker, R.: An Application of the Principle of Maximum Information Preservation to Linear Systems. In: Touretzky, D. (ed.) Advances in Neural Information Processing Systems, vol. 1. Morgan-Kaufmann, San Francisco (1989)Google Scholar
- 5.Agakov, F.V., Barber, D.: Variational Information Maximization for Neural Coding. In: International Conference on Neural Information Processing. Springer, Heidelberg (2004)Google Scholar
- 6.Agakov, F.V.: Variational Information Maximization in Stochastic Environments. PhD thesis, School of Informatics, University of Edinburgh (2005)Google Scholar
- 9.Linsker, R.: Deriving Receptive Fields Using an Optimal Encoding Criterion. In: Hanson, S., Cowan, J., Gilese, L. (eds.) Advances in Neural Information Processing Systems, vol. 5. Morgan Kaufmann, San Francisco (1993)Google Scholar
- 10.Neal, R.M., Hinton, G.E.: A View of the EM Algorithm That Justifies Incremental, Sparse, and Other Variants. In: Jordan, M. (ed.) Learning in Graphical Models. Kluwer Academic, Dordrecht (1998)Google Scholar
- 12.Arimoto, S.: An Algorithm for computing the capacity of arbitrary discrete memoryless channels. IT 18 (1972)Google Scholar
- 13.Blahut, R.: Computation of channel capacity and rate-distortion functions. IT 18 (1972)Google Scholar
- 14.Jacobs, R.A., Jordan, M.I., Nowlan, S.J., Hinton, G.E.: Adaptive Mixtures of Local Experts. Neural Computation 3 (1991)Google Scholar
- 15.Tishby, N., Pereira, F.C., Bialek, W.: The information bottleneck method. In: Proc. of the 37-th Annual Allerton Conference on Communication, Control and Computing, pp. 368–377 (1999)Google Scholar
- 16.LeCun, Y., Cortes, C.: The MNIST database of handwritten digits (1998)Google Scholar
- 17.Cardoso, J.F.: Infomax and maximum likelihood for blind source separation. IEEE Signal Processing Letters 4 (1997)Google Scholar