English

Optimal Sketching for Trace Estimation

Data Structures and Algorithms 2021-11-02 v1 Numerical Analysis Numerical Analysis

Abstract

Matrix trace estimation is ubiquitous in machine learning applications and has traditionally relied on Hutchinson's method, which requires O(log(1/δ)/ϵ2)O(\log(1/\delta)/\epsilon^2) matrix-vector product queries to achieve a (1±ϵ)(1 \pm \epsilon)-multiplicative approximation to tr(A)\text{tr}(A) with failure probability δ\delta on positive-semidefinite input matrices AA. Recently, the Hutch++ algorithm was proposed, which reduces the number of matrix-vector queries from O(1/ϵ2)O(1/\epsilon^2) to the optimal O(1/ϵ)O(1/\epsilon), and the algorithm succeeds with constant probability. However, in the high probability setting, the non-adaptive Hutch++ algorithm suffers an extra O(log(1/δ))O(\sqrt{\log(1/\delta)}) multiplicative factor in its query complexity. Non-adaptive methods are important, as they correspond to sketching algorithms, which are mergeable, highly parallelizable, and provide low-memory streaming algorithms as well as low-communication distributed protocols. In this work, we close the gap between non-adaptive and adaptive algorithms, showing that even non-adaptive algorithms can achieve O(log(1/δ)/ϵ+log(1/δ))O(\sqrt{\log(1/\delta)}/\epsilon + \log(1/\delta)) matrix-vector products. In addition, we prove matching lower bounds demonstrating that, up to a loglog(1/δ)\log \log(1/\delta) factor, no further improvement in the dependence on δ\delta or ϵ\epsilon is possible by any non-adaptive algorithm. Finally, our experiments demonstrate the superior performance of our sketch over the adaptive Hutch++ algorithm, which is less parallelizable, as well as over the non-adaptive Hutchinson's method.

Keywords

Cite

@article{arxiv.2111.00664,
  title  = {Optimal Sketching for Trace Estimation},
  author = {Shuli Jiang and Hai Pham and David P. Woodruff and Qiuyi and Zhang},
  journal= {arXiv preprint arXiv:2111.00664},
  year   = {2021}
}

Comments

31 pages, 5 figures. Proceedings of the 35th Conference on Neural Information Processing Systems (NeurIPS 2021), Sydney, Australia

R2 v1 2026-06-24T07:20:11.721Z