Abstract
The Kaczmarz algorithm is an iterative method for reconstructing a signal x∈ℝd from an overcomplete collection of linear measurements y n =〈x,φ n 〉, n≥1. We prove quantitative bounds on the rate of almost sure exponential convergence in the Kaczmarz algorithm for suitable classes of random measurement vectors \(\{\varphi_{n}\}_{n=1}^{\infty} \subset {\mathbb {R}}^{d}\). Refined convergence results are given for the special case when each φ n has i.i.d. Gaussian entries and, more generally, when each φ n /∥φ n ∥ is uniformly distributed on \(\mathbb{S}^{d-1}\). This work on almost sure convergence complements the mean squared error analysis of Strohmer and Vershynin for randomized versions of the Kaczmarz algorithm.
Similar content being viewed by others
References
Editorial comment on the two following letters to the editor: concerning Kaczmarz algorithms. J. Fourier Anal. Appl. 15(4) 429–430 (2009)
Censor, Y., Herman, G., Jiang, M.: A note on the behavior of the randomized Kaczmarz algorithm of Strohmer and Vershynin. J. Fourier Anal. Appl. 15(4), 431–436 (2009)
Ehler, M., Okoudjou, K.: Probabilistic frames: an overview. In: Casazza, P., Kutyniok, G. (eds.) Finite Frames: Theory and Applications. Birkhäuser, to appear
Eldar, Y., Needell, D.: Acceleration of randomized Kaczmarz method via the Johnson-Lindenstrauss lemma. Numer. Algorithms 58(2), 163–177 (2011)
Haller, R., Szwarc, R.: Kaczmarz algorithm in Hilbert space. Studia Math. 169(2), 123–132 (2005)
Hwang, C.-R., Sheu, S.-J.: On the geometrical convergence of the Gibbs sampler in R d∗. J. Multivar. Anal. 66, 22–37 (1998)
Kaczmarz, S.: Angenäherte Auflösung von Systemen linearer Gleichungen. Bull. Int. Acad. Pol. Sci. A, 355–357 (1937). English translation in: Approximate solution of systems of linear equations. Int. J. Control 57, 1269–1271 (1993)
Kwapień, S., Mycielski, J.: On the Kaczmarz algorithm of approximation in infinite-dimensional spaces. Studia Math. 148(1), 75–86 (2001)
Leventhal, D., Lewis, A.S.: Randomized methods for linear constraints: convergence rates and conditioning. Math. Oper. Res. 35(3), 641–654 (2010)
Milnor, J.: Hyperbolic geometry: the first 150 years. Bull. Am. Math. Soc. 6(1), 9–24 (1982)
Natterer, F.: The Mathematics of Computerized Tomography. Wiley, New York (1986)
Needell, D.: Randomized Kaczmarz solver for noisy linear systems. BIT Numer. Math. 50(2), 395–403 (2010)
Powell, A.M.: Mean squared error bounds for the Rangan-Goyal soft thresholding algorithm. Appl. Comput. Harmon. Anal. 29(3), 251–271 (2010)
Rangan, S., Goyal, V.K.: Recursive consistent estimation with bounded noise. IEEE Trans. Inf. Theory 47(1), 457–464 (2001)
Rudin, W.: Real and Complex Analysis, 3rd edn. McGraw-Hill, New York (1987)
Saff, E., Totik, V.: Logarithmic Potentials with External Fields, 1st edn. Springer, New York (1997)
Sezan, K.M., Stark, H.: Applications of convex projection theory to image recovery in tomography and related areas. In: Stark, H. (ed.) Image Recovery: Theory and application, pp. 415–462. Academic Press, San Diego (1987)
Shiryaev, A.N.: Probability, 2nd edn. Springer, New York (1989)
Sra, S., Tropp, J.: Row-action methods for compressed sensing. In: Proceedings of the 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Toulouse, vol. 3, pp. 868–871 (2006)
Strohmer, T., Vershynin, R.: A randomized solver for linear systems with exponential convergence. In: Approximation, Randomization and Combinatorial Optimization. Lecture Notes in Computer Science, vol. 4110, pp. 499–507. Springer, Berlin (2006)
Strohmer, T., Vershynin, R.: A randomized Kaczmarz algorithm with exponential convergence. J. Fourier Anal. Appl. 15(2), 262–278 (2009)
Strohmer, T., Vershynin, R.: Comments on the randomized Kaczmarz method. J. Fourier Anal. Appl. 15(4), 437–440 (2009)
Szwarc, R.: Kaczmarz algorithm in Hilbert space and tight frames. Appl. Comput. Harmon. Anal. 22(3), 382–385 (2007)
Acknowledgements
A. Powell was supported in part by NSF DMS Grant 0811086, and gratefully acknowledges the Academia Sinica Institute of Mathematics (Taipei, Taiwan) for its hospitality and support. He especially thanks Professors Yunshyong Chou, Chii-Ruey Hwang, and Shuenn-Jyi Sheu for valuable discussions related to the material.
The authors thank the anonymous reviewers for their helpful comments and especially for pointing out the reference [9].
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Roman Vershynin.
Rights and permissions
About this article
Cite this article
Chen, X., Powell, A.M. Almost Sure Convergence of the Kaczmarz Algorithm with Random Measurements. J Fourier Anal Appl 18, 1195–1214 (2012). https://doi.org/10.1007/s00041-012-9237-2
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00041-012-9237-2