English

An Improved Classical Singular Value Transformation for Quantum Machine Learning

Quantum Physics 2024-07-30 v4 Data Structures and Algorithms

Abstract

We study quantum speedups in quantum machine learning (QML) by analyzing the quantum singular value transformation (QSVT) framework. QSVT, introduced by [GSLW, STOC'19, arXiv:1806.01838], unifies all major types of quantum speedup; in particular, a wide variety of QML proposals are applications of QSVT on low-rank classical data. We challenge these proposals by providing a classical algorithm that matches the performance of QSVT in this regime up to a small polynomial overhead. We show that, given a matrix ACm×nA \in \mathbb{C}^{m\times n}, a vector bCnb \in \mathbb{C}^{n}, a bounded degree-dd polynomial pp, and linear-time pre-processing, we can output a description of a vector vv such that vp(A)bεb\|v - p(A) b\| \leq \varepsilon\|b\| in O~(d11AF4/(ε2A4))\widetilde{\mathcal{O}}(d^{11} \|A\|_{\mathrm{F}}^4 / (\varepsilon^2 \|A\|^4 )) time. This improves upon the best known classical algorithm [CGLLTW, STOC'20, arXiv:1910.06151], which requires O~(d22AF6/(ε6A6))\widetilde{\mathcal{O}}(d^{22} \|A\|_{\mathrm{F}}^6 /(\varepsilon^6 \|A\|^6 ) ) time, and narrows the gap with QSVT, which, after linear-time pre-processing to load input into a quantum-accessible memory, can estimate the magnitude of an entry p(A)bp(A)b to εb\varepsilon\|b\| error in O~(dAF/(εA))\widetilde{\mathcal{O}}(d\|A\|_{\mathrm{F}}/(\varepsilon \|A\|)) time. Our key insight is to combine the Clenshaw recurrence, an iterative method for computing matrix polynomials, with sketching techniques to simulate QSVT classically. We introduce several new classical techniques in this work, including (a) a non-oblivious matrix sketch for approximately preserving bi-linear forms, (b) a new stability analysis for the Clenshaw recurrence, and (c) a new technique to bound arithmetic progressions of the coefficients appearing in the Chebyshev series expansion of bounded functions, each of which may be of independent interest.

Keywords

Cite

@article{arxiv.2303.01492,
  title  = {An Improved Classical Singular Value Transformation for Quantum Machine Learning},
  author = {Ainesh Bakshi and Ewin Tang},
  journal= {arXiv preprint arXiv:2303.01492},
  year   = {2024}
}

Comments

63 pages, v4: minor edits for referee comments, v3: fixed bug, runtime exponent now 11 instead of 9; v2: revised abstract to clarify results

R2 v1 2026-06-28T08:57:57.418Z