New Null Space Results and Recovery Thresholds for Matrix Rank Minimization

Nuclear norm minimization (NNM) has recently gained significant attention for its use in rank minimization problems. Similar to compressed sensing, using null space characterizations, recovery thresholds for NNM have been studied in. However simulations show that the thresholds are far from optimal, especially in the low rank region. In this paper we apply the recent analysis of Stojnic for compressed sensing to the null space conditions of NNM. The resulting thresholds are significantly better and in particular our weak threshold appears to match with simulation results. Further our curves suggest for any rank growing linearly with matrix size n we need only three times of oversampling (the model complexity) for weak recovery. Similar to we analyze the conditions for weak, sectional and strong thresholds. Additionally a separate analysis is given for special case of positive semidefinite matrices. We conclude by discussing simulation results and future research directions.

[1]  V. Marčenko,et al.  DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES , 1967 .

[2]  L. Mirsky.  A trace inequality of John von Neumann , 1975 .

[3]  J. Kuelbs.  Probability on Banach spaces , 1978 .

[4]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[5]  Y. Gordon.  Some inequalities for Gaussian processes and applications , 1985 .

[6]  V. Girko.  An Introduction to Statistical Analysis of Random Arrays , 1998 .

[7]  Chi-Kwong Li,et al.  The Lidskii-Mirsky-Wielandt theorem – additive and multiplicative versions , 1999, Numerische Mathematik.

[8]  David A. Gregory,et al.  Inertia and biclique decompositions of joins of graphs , 2003, J. Comb. Theory, Ser. B.

[9]  D. Donoho,et al.  Neighborliness of randomly projected simplices in high dimensions. , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[10]  D. Donoho,et al.  Thresholds for the Recovery of Sparse Solutions via L1 Minimization , 2006, 2006 40th Annual Conference on Information Sciences and Systems.

[11]  E. Candès.  The restricted isometry property and its implications for compressed sensing , 2008 .

[12]  Weiyu Xu,et al.  Compressed sensing - probabilistic analysis of a null-space characterization , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[13]  Mihailo Stojnic,et al.  Various thresholds for ℓ1-optimization in compressed sensing , 2009, ArXiv.

[14]  M. Stojnic.  Various thresholds for $\ell_1$-optimization in compressed sensing , 2009 .

[15]  Pablo A. Parrilo,et al.  Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization , 2007, SIAM Rev..

[16]  Ao Tang,et al.  On the uniqueness of positive semidefinite matrix solution under compressed observations , 2010, 2010 IEEE International Symposium on Information Theory.

[17]  Emmanuel J. Candès,et al.  Tight oracle bounds for low-rank matrix recovery from a minimal number of random measurements , 2010, ArXiv.

[18]  Babak Hassibi,et al.  Improved thresholds for rank minimization , 2011, 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[19]  Weiyu Xu,et al.  Null space conditions and thresholds for rank minimization , 2011, Math. Program..

[20]  Emmanuel J. Candès,et al.  Exact Matrix Completion via Convex Optimization , 2008, Found. Comput. Math..