Non-Convex Rank Minimization via an Empirical Bayesian Approach

In many applications that require matrix solutions of minimal rank, the underlying cost function is non-convex leading to an intractable, NP-hard optimization problem. Consequently, the convex nuclear norm is frequently used as a surrogate penalty term for matrix rank. The problem is that in many practical scenarios there is no longer any guarantee that we can correctly estimate generative low-rank matrices of interest, theoretical special cases notwithstanding. Consequently, this paper proposes an alternative empirical Bayesian procedure build upon a variational approximation that, unlike the nuclear norm, retains the same globally minimizing point estimate as the rank function under many useful constraints. However, locally minimizing solutions are largely smoothed away via marginalization, allowing the algorithm to succeed when standard convex relaxations completely fail. While the proposed methodology is generally applicable to a wide range of low-rank applications, we focus our attention on the robust principal component analysis problem (RPCA), which involves estimating an unknown low-rank matrix with unknown sparse corruptions. Theoretical and empirical evidence are presented to show that our method is potentially superior to related MAP-based approaches, for which the convex principle component pursuit (PCP) algorithm (Candes et al., 2011) can be viewed as a special case.

[1]  Yi Ma,et al.  The Augmented Lagrange Multiplier Method for Exact Recovery of Corrupted Low-Rank Matrices , 2010, Journal of structural biology.

[2]  M. Fazel,et al.  Iterative reweighted least squares for matrix rank minimization , 2010, 2010 48th Annual Allerton Conference on Communication, Control, and Computing (Allerton).

[3]  Yi Ma,et al.  Robust principal component analysis? , 2009, JACM.

[4]  G. Sapiro,et al.  A collaborative framework for 3D alignment and classification of heterogeneous subvolumes in cryo-electron tomography. , 2013, Journal of structural biology.

[5]  Pablo A. Parrilo,et al.  Rank-Sparsity Incoherence for Matrix Decomposition , 2009, SIAM J. Optim..

[6]  Robert L. Cook,et al.  A Reflectance Model for Computer Graphics , 1987, TOGS.

[7]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[8]  Emmanuel J. Candès,et al.  Exact Matrix Completion via Convex Optimization , 2009, Found. Comput. Math..

[9]  George Eastman House,et al.  Sparse Bayesian Learning and the Relevance Vector Machine , 2001 .

[10]  J. Berger Statistical Decision Theory and Bayesian Analysis , 1988 .

[11]  David P. Wipf,et al.  Iterative Reweighted 1 and 2 Methods for Finding Sparse Solutions , 2010, IEEE J. Sel. Top. Signal Process..

[12]  Aggelos K. Katsaggelos,et al.  Sparse Bayesian Methods for Low-Rank Matrix Estimation , 2011, IEEE Transactions on Signal Processing.

[13]  Lawrence Carin,et al.  Bayesian Robust Principal Component Analysis , 2011, IEEE Transactions on Image Processing.

[14]  Bhaskar D. Rao,et al.  Latent Variable Bayesian Models for Promoting Sparsity , 2011, IEEE Transactions on Information Theory.

[15]  Robert J. Woodham,et al.  Photometric method for determining surface orientation from multiple images , 1980 .

[16]  Michael I. Jordan,et al.  An Introduction to Variational Methods for Graphical Models , 1999, Machine-mediated learning.

[17]  Hagai Attias,et al.  A Variational Bayesian Framework for Graphical Models , 1999 .

[18]  Yongtian Wang,et al.  Robust Photometric Stereo via Low-Rank Matrix Completion and Recovery , 2010, ACCV.