The phase transition of matrix recovery from Gaussian measurements matches the minimax MSE of matrix denoising

Proceedings of the National Academy of Sciences of the United States of America

Volumn 110, Issue 21, 2013, Pages 8405-8410

  Donoho, David L ^a   Gavish, Matan ^a   Montanari, Andrea ^a  

^a STANFORD UNIVERSITY   (United States)

Author keywords

Compressed sensing; Matrix completion

Indexed keywords

ARTICLE; MEASUREMENT; NOISE; PHASE TRANSITION; PRIORITY JOURNAL;

EID: 84878157685     PISSN: 00278424     EISSN: 10916490     Source Type: Journal    
DOI: 10.1073/pnas.1306110110     Document Type: Article

References (40)

1. Recht B, Fazel M, Parrilo PA (2010) Guaranteed minimum-rank solutions of linear matrix equations via nuclear norm minimization. SIAM Rev 52(3): 471-501.
2. Candès EJ, Recht B (2008) Exact matrix completion via convex optimization. Found Comput Math 9(6): 717-772.
3. Gross D (2011) Recovering low-rank matrices from few coefficients in any basis. IEEE Trans Inf Theory 57(3): 1548-1566.
4. Candès EJ, Li X, Ma Y, Wright J (2009) Robust principal component analysis? JACM 58(3), 10.1145/1970392.1970395.
5. Gross D, Liu YK, Flammia ST, Becker S, Eisert J (2010) Quantum state tomography via compressed sensing. Phys Rev Lett 105(15):150401.
6. Keshavan RH, Montanari A, Oh S (2010) Matrix completion from noisy entries. The Journal of Machine Learning 11: 2057-2078.
7. Recht B, Xu W, Hassibi B (2010) Null space conditions and thresholds for rank minimization. Math Program 127(1): 175-202.
8. Recht B, Xu W, Hassibi B (2008) Necessary and sufficient conditions for success of the nuclear norm heuristic for rank minimization. Proceedings of the 47th IEEE Conference on Decision and Control (IEEE, New York), pp 3065-3070.
9. Tanner J, Wei K (2012) Normalized iterative hard thresholding for matrix completion. Available at http://people.maths.ox.ac.uk/tanner/papers/TaWei\-NIHT. pdf. Accessed April 15, 2013.
10. Oymak S, Hassibi B (2010) New null space results and recovery thresholds for matrix rank minimization. Available at http://arxiv.org/abs/1011.6326. Accessed April 15, 2013.
11. Donoho DL, Gavish M (2013) Minimax Risk of Matrix Denoising by Singular Value Thresholding (Stanford Univ Dept. of Statistics, Stanford, CA). Available at http://arxiv. org/abs/1304.2085. Accessed April 15, 2013. Tech Rep 2013-03.
12. Fazel M, Hindi H, Boyd SP (2001) A rank minimization heuristic with application to minimum order system approximation. Proceedings of the 2001 American Control Conference (IEEE, New York), pp 4734-4739.
13. Fazel M (2002) Matrix rank minimization and applications. PhD thesis (Stanford Univ, Stanford, CA).
14. Donoho DL, Tanner J (2009) Observed universality of phase transitions in highdimensional geometry, with implications for modern data analysis and signal processing. Philos Transact A Math Phys Eng Sci 367(1906): 4273-4293.
15. Monajemi H, Jafarpour S, Gavish M, Donoho DL; Stat 330/CME 362 Collaboration (2013) Deterministic matrices matching the compressed sensing phase transitions of Gaussian random matrices. Proc Natl Acad Sci USA 110(4): 1181-1186.
16. Kalish LA (1990) Efficient design for estimation of median lethal dose and quantal dose-response curves. Biometrics 46(3): 737-748.
17. Grant M, Boyd SP (2010) CVX: Matlab Software for Disciplined Convex Programming (CVX Research Inc.). Available at http://cxvr.com/cvx. Accessed April 15, 2013.
18. Toh KC, Todd MJ, Tutuncu RH (1999) SDPT3 - A Matlab software package for semidefinite programming. Optimization Methods Software.
19. Sturm JF (1999) Using SeDuMi 1.02, a MATLAB toolbox for optimization over symmetric cones. Optimization Methods Software. Available at http://www.tandfonline. com/doi/abs/10.1080/10556789908805766. Accessed April 15, 2013.
20. Andersen D, Dahl J, Vandenberghe L (2012) CVXOPT: Python software for convex optimization. Available at http://abel.ee.ucla.edu/cvxopt. Accessed April 15, 2013.
21. Donoho DL, Gavish M, Montanari A (2013) Data for the article 'The phase transition of matrix recovery from gaussian measurements matches the minimax MSE of matrix denoising.' Available at http://purl.stanford.edu/tz124hw0000. Accessed February 15, 2013.
22. Donoho DL, Tanner J (2005) Neighborliness of randomly projected simplices in high dimensions. Proc Natl Acad Sci USA 102(27): 9452-9457.
23. Donoho DL (2005) High-dimensional centrally symmetric polytopes with neighborliness proportional to dimension. Discrete Comput Geom 35(4): 617-652.
24. Donoho DL, Tanner J (2009) Counting faces of randomly projected polytopes when the projection radically lowers dimension. J Am Math Soc 22(1): 1-53.
25. Donoho DL, Tanner J (2010) Precise undersampling theorems. Proc IEEE 98(6): 913-924.
26. Donoho DL, Maleki A, Montanari A (2009) Message-passing algorithms for compressed sensing. Proc Natl Acad Sci USA 106(45): 18914-18919.
27. BayatiM, Montanari A (2011) The dynamics of message passing on dense graphs, with applications to compressed sensing. IEEE Trans Inf Theory 57(2): 764-785.
28. Bayati M, Montanari A (2012) The LASSO risk for Gaussian matrices. IEEE Trans Inf Theory 58(4): 1997-2017.
29. Vila J, Schniter P (2011) Expectation-maximization Bernoulli-Gaussian approximate message passing. Conference Record of the Forty-Fifth ASILOMAR Conference on Signals, Systems and Computers (IEEE, New York), pp 799-803.
30. Rangan S (2011) Generalized approximate message passing for estimation with random linear mixing. 2011 IEEE International Symposium on Information Theory Proceedings (IEEE, New York), pp 2168-2172.
31. Bayati M, Lelarge M, Montanari A (2012) Universality in polytope phase transitions and message passing algorithms. Available at http://arxiv.org/abs/ 1207.7321. Accessed April 15, 2013.
32. Donoho DL, Johnstone IM, Montanari A (2011) Accurate prediction of phase transitions in compressed sensing via a connection to minimax denoising. Available at http://arxiv.org/abs/1111.1041. Accessed April 15, 2013.
33. Oymak S, Mohan K, Fazel M, Hassibi B (2011) A simplified approach to recovery conditions for low rank matrices. 2011 IEEE International Symposium on Information Theory Proceedings (IEEE, New York), pp 2318-2322.
34. Candès EJ, Recht B (2011) Simple bounds for recovering low-complexity models. Math Program, 1-11. Available at www.springerlink.com/ index/655751h7568x3621.pdf. Accessed April 15, 2013.
35. Gordon Y (1988) On Milman's inequality and random subspaces which escape through a mesh in RN. Lecture Notes in Mathematics (Springer, Berlin), Vol 1317, pp 84-106.
36. Stojnic M (2009) Various thresholds for l1-optimization in compressed sensing. Available at http://arxiv.org/abs/0907.3666. Accessed April 15, 2013.
37. Oymak S, Hassibi B (2012) On a relation between the minimax denoising and the phase transitions of convex functions. Available at www.its.caltech.edu/ ~soymak/ Relation.pdf. Accessed April 15, 2013.
38. Donoho DL, Gavish M (2013) Companion website for the article 'The phase transition of matrix recovery from Gaussian measurements matches the minimax MSE of matrix denoising. Available at www.runmycode.org/CompanionSite/Site265. Accessed February 15, 2013.
39. Marcenko V, Pastur L (1967) Distribution of eigenvalues for some sets of random matrices. Mathematics USSR Sbornik 1(4): 457-483.
40. Amelunxen D, Lotz M, McCoy MB, Tropp JA (2013) Living on the edge: A geometric theory of phase transitions in convex optimization. Available at http://arxiv.org/abs/ 1303.6672. Accessed April 15, 2013.