Convergence of Fixed-Point Continuation Algorithms for Matrix Rank Minimization
- 688 Downloads
The matrix rank minimization problem has applications in many fields, such as system identification, optimal control, low-dimensional embedding, etc. As this problem is NP-hard in general, its convex relaxation, the nuclear norm minimization problem, is often solved instead. Recently, Ma, Goldfarb and Chen proposed a fixed-point continuation algorithm for solving the nuclear norm minimization problem (Math. Program., doi: 10.1007/s10107-009-0306-5, 2009). By incorporating an approximate singular value decomposition technique in this algorithm, the solution to the matrix rank minimization problem is usually obtained. In this paper, we study the convergence/recoverability properties of the fixed-point continuation algorithm and its variants for matrix rank minimization. Heuristics for determining the rank of the matrix when its true rank is not known are also proposed. Some of these algorithms are closely related to greedy algorithms in compressed sensing. Numerical results for these algorithms for solving affinely constrained matrix rank minimization problems are reported.
KeywordsMatrix rank minimization Matrix completion Greedy algorithm Fixed-point method Restricted isometry property Singular value decomposition
Mathematics Subject Classification (2000)90C59 15B52 15A18
Unable to display preview. Download preview PDF.
- 3.J.M. Borwein, A.S. Lewis, Convex Analysis and Nonlinear Optimization (Springer, Berlin, 2003). Google Scholar
- 5.E.J. Candès, Y. Plan, Matrix completion with noise, Proc. IEEE (2009). Google Scholar
- 7.E.J. Candès, J. Romberg, ℓ 1-MAGIC: Recovery of sparse signals via convex programming, Tech. rep., Caltech, 2005. Google Scholar
- 10.Rice compressed sensing website. http://dsp.rice.edu/cs.
- 14.D. Donoho, Y. Tsaig, I. Drori, J.-C. Starck, Sparse solution of underdetermined linear equations by stagewise orthogonal matching pursuit, Tech. rep., Stanford University, 2006. Google Scholar
- 16.M. Fazel, H. Hindi, S. Boyd, A rank minimization heuristic with application to minimum order system approximation, in Proceedings of the American Control Conference, vol. 6 (2001), pp. 4734–4739. Google Scholar
- 17.M. Fazel, H. Hindi, S. Boyd, Log-det heuristic for matrix rank minimization with applications to Hankel and Euclidean distance matrices, in Proceedings of the American Control Conference (2003), pp. 2156–2162. Google Scholar
- 18.M. Fazel, H. Hindi, S. Boyd, Rank minimization and applications in system theory, in American Control Conference (2004), pp. 3273–3278. Google Scholar
- 20.L.E. Ghaoui, P. Gahinet, Rank minimization under LMI constraints: A framework for output feedback problems, in Proceedings of the European Control Conference (1993). Google Scholar
- 23.R.H. Keshavan, A. Montanari, S. Oh, Matrix completion from noisy entries (2009). arXiv:0906.2027.
- 26.R.M. Larsen, PROPACK—software for large and sparse SVD calculations, available from http://sun.stanford.edu/~rmunk/PROPACK.
- 27.K. Lee, Y. Bresler, ADMIRA: atomic decomposition for minimum rank approximation (2009). arXiv:0905.0044.
- 28.K. Lee, Y. Bresler, Efficient and guaranteed rank minimization by atomic decomposition (2009). arXiv:0901.1898v1.
- 29.K. Lee, Y. Bresler, Guaranteed minimum rank approximation from linear observations by nuclear norm minimization with an ellipsoidal constraint (2009). arXiv:0903.4742.
- 31.Y. Liu, D. Sun, K.-C. Toh, An implementable proximal point algorithmic framework for nuclear norm minimization, Preprint, National University of Singapore, 2009. Google Scholar
- 34.R. Meka, P. Jain, I.S. Dhillon, Guaranteed rank minimization via singular value projection (2009). arXiv:0909.5457.
- 37.Netfix prize website. http://www.netflixprize.com/.
- 40.N. Srebro, Learning with Matrix Factorizations. PhD thesis, Massachusetts Institute of Technology, 2004. Google Scholar
- 41.N. Srebro, T. Jaakkola, Weighted low-rank approximations, in Proceedings of the Twentieth International Conference on Machine Learning (ICML-2003) (2003). Google Scholar