paper-with-me

홈 › Papers

Hutch++: Optimal Stochastic Trace Estimation

2020-10-19 · Raphael A. Meyer, Cameron Musco, Christopher Musco, David P. Woodruff

We study the problem of estimating the trace of a matrix $A$ that can only be accessed through matrix-vector multiplication. We introduce a new randomized algorithm, Hutch++, which computes a $(1 \pm \epsilon)$ approximation to $tr(A)$ for any positive semidefinite (PSD) $A$ using just $O(1/\epsilon)$ matrix-vector products. This improves on the ubiquitous Hutchinson's estimator, which requires $O(1/\epsilon^2)$ matrix-vector products. Our approach is based on a simple technique for reducing the variance of Hutchinson's estimator using a low-rank approximation step, and is easy to implement and analyze. Moreover, we prove that, up to a logarithmic factor, the complexity of Hutch++ is optimal amongst all matrix-vector query algorithms, even when queries can be chosen adaptively. We show that it significantly outperforms Hutchinson's method in experiments. While our theory mainly requires $A$ to be positive semidefinite, we provide generalized guarantees for general square matrices, and show empirical gains in such applications.

📄 PDF Abstract BibTeX arXiv:2010.09649

Code (1)

RaphaelArkadyMeyerNYU/HutchPlusPlus 공식 구현

Similar Papers 제목 키워드 기반

Optimal Stochastic Trace Estimation in Generative Modeling

2025-02-26 · Xinyang Liu, Hengrong Du, Wei Deng, Ruqi Zhang

Hutchinson estimators are widely employed in training divergence-based likelihoods for diffusion models to ensure optimal transport (OT) properties. However, this estimator often suffers from high variance and scalabilit…

Image GenerationTime Series

BOLT: Block-Orthonormal Lanczos for Trace estimation of matrix functions

2025-05-18 · Kingsley Yeon, Promit Ghosal, Mihai Anitescu

Efficient matrix trace estimation is essential for scalable computation of log-determinants, matrix norms, and distributional divergences. In many large-scale applications, the matrices involved are too large to store or…

subspace methods

Estimation of matrix trace using machine learning

2016-06-16 · Boram Yoon

We present a new trace estimator of the matrix whose explicit form is not given but its matrix multiplication to a vector is available. The form of the estimator is similar to the Hutchison stochastic trace estimator, bu…

BIG-bench Machine LearningForm

Efficient DP-SGD for LLMs with Randomized Clipping

2026-05-24 · Enayat Ullah, Sai Aparna Aketi, Devansh Gupta, Huanyu Zhang 외 arxiv

Large language models (LLMs) are trained on vast datasets that may contain sensitive information. Differential privacy (DP), the de facto standard for formal privacy guarantees, provides a principled framework for traini…

Question Answering

Stochastic trace estimation with tensor train random vectors

2026-06-14 · Zvonimir Bujanović, Daniel Kressner, Hrvoje Olić arxiv

Stochastic trace estimation is a standard tool for approximating the trace of a large-scale matrix available only through matrix-vector products. However, in tensor-structured settings, unstructured Gaussian or Rademache…