Abstract
This paper develops the Bernstein tensor concentration inequality for random tensors of general order, based on the use of Einstein products for tensors. This establishes a strong link between these and matrices, which in turn allows exploitation of existing results for the latter. An interesting application to sample estimators of high-order moments is presented as an illustration.
Similar content being viewed by others
References
Aclioptas D, McSherry F. Fast computation of low rank matrix approximations. In: Proc 33rd Ann ACM Symposium on Theory of Computing. 2001, 611-618
Ahlswede R, Winter A. Strong converse for identification via quantum channels. IEEE Trans Inform Theory, 2002, 48(3): 569–579
Bellavia S, Gurioli G, Morini B. Theoretical study of an adaptive cubic regularization method with dynamic inexact Hessian information. arXiv: 1808.06239
Bellavia S, Gurioli G, Morini B, Toint Ph L. Deterministic and stochastic inexact regularization algorithms for nonconvex optimization with optimal complexity. arXiv: 1811.03831
Bellavia S, Krejic N, Krklec Jerinkic N. Subsampled inexact Newton methods for minimizing large sums of convex function. arXiv: 1811.05730
Brazell M, Li N, Navasca C, Tamon Ch. Solving multilinear systems via tensor inversion. SIAM J Matrix Anal Appl, 2013, 34(2): 542–570
Cartwright D, Sturmfels B. The number of eigenvalues of a tensor. Linear Algebra Appl, 2013, 438(2): 942–952
Chen X, Jiang B, Lin T, Zhang S. On adaptive cubic regularization Newton’s methods for convex optimization via random sampling. arXiv: 1802.05426
Donoho D L. Compressed sensing. IEEE Trans Inform Theory, 2006, 52(4): 1289–1306
Einstein A. The foundation of the general theory of relativity. In: The Collected Papers of Albert Einstein, Vol 6. Princeton: Princeton Univ Press, 1997, 146–200
Forrester P J. Log-Gases and Random Matrices. London Math Soc Monogr Ser, 34. Princeton: Princeton Univ Press, 2010
Frieze A, Kannan R, Vempala S. Fast Monte-Carlo algorithms for finding low-rank approximations. In: Proc 39th Ann Symposium on Foundations of Computer Science. 1998, 370–378
Kohier J M. Lucchi A. Sub-sample cubic regularization for non-convex optimization. In: Proc 34th International Conference on Machine Learning, Sydney, 2017. arXiv: 1705.05933v3
Lai W M, Rubin D H, Krempl E. Introduction to Continuum Mechanics. Oxford: Butterworth-Heinemann, 2009
Luo Z, Qi L, Ye Y. Linear operators and positive semidefiniteness of symmetric tensor spaces. Sci China Math, 2015, 58(1): 197–212
Miao Y, Qi L, Wei Y. Generalized tensor function via the tensor singular value decomposition based on the T-product. Linear Algebra Appl, 2020, 590: 258–303
Qi L. Eigenvalues of a real supersymmetric tensor. J Symbolic Comput, 2005, 40(6): 1302–1324
Qi L, Luo Z. Tensor Analysis: Spectral Theory and Special Tensors. Philadelphia: SIAM, 2017
Qi L, Yu G, Wu E X. Higher order positive semidefinite diffusion tensor imaging. SIAM J Imaging Sci, 2010, 3(3): 416–433
Sun L, Zheng B, Bu C, Wei Y. Moore-Penrose inverse of tensors via Einstein product. Linear Multilinear Algebra, 2016, 64(4): 686–698
Thomson W. Elements of a mathematical theory of elasticity. Proc Roy Soc Lond, 1856, 8: 85–87
Tropp J. An Introduction to Matrix Concentration Inequalities. Found Trends Machine Learning, Number 8: 1–2. Boston: Now Publ, 2015
Williams C K I, Seeger N. Using the Nyström method to speed up kernel machines. In: Advances in Neural Information Processing Systems 13. 2001, 682–688
Wishart J. The generalized product moment distribution in samples from a multivariate normal population. Biometrika, 1928, 20A(1–2): 32–52
Xu P, Roosta-Khorasani F, Mahoney M W. Newton-type methods for non-convex optimization under inexact Hessian information. arXiv: 1708.07164v3
Xu P, Roosta-Khorasani F, Mahoney M W. Second-order optimization for non-convex machine learning: an empirical study. arXiv: 1708.07827v2
Acknowledgements
This work was partially supported by the National Natural Science Foundation of China (Grant No. 11771038) and the Beijing Natural Science Foundation (Grant No. Z190002), and the Hong Kong Research Grant Council (Grant Nos. PolyU 15300715, 15301716, 15300717). The third author gratefully acknowledges the support from Prof. Xiaojun Chen in Hong Kong Polytechnic University for the visit during which this research was initiated.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Luo, Z., Qi, L. & Toint, P.L. Tensor Bernstein concentration inequalities with an application to sample estimators for high-order moments. Front. Math. China 15, 367–384 (2020). https://doi.org/10.1007/s11464-020-0830-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11464-020-0830-4