An Improved Classical Singular Value Transformation for Quantum Machine Learning
Ainesh Bakshi, Ewin Tang
TL;DR
This work shows that for low-rank classical data, a classical approach based on the Clenshaw recurrence combined with novel sketching and bilinear sampling techniques can approximate $p(A)b$ as effectively as QSVT, up to polylog and poly-factor overheads. By constructing dimension-free sketches and carefully analyzing finite-precision stability, the authors achieve running times with improved dependencies on the polynomial degree $d$, Frobenius norm $ norm{A}$, and error $varepsilon$ relative to prior dequantization work. The resulting framework yields faster quantum-inspired algorithms for regression, recommendation systems, and Hamiltonian simulation, illustrating a concrete quantum-classical gap in the low-rank regime and highlighting when quantum speedups may be realized. The techniques—including BEST sketches, AMP-based matrix product sketches, and refined Chebyshev-sum bounds—also contribute broadly to randomized numerical linear algebra and the study of polynomial matrix functions. Overall, the paper advances both theoretical understanding and practical tools for dequantizing QML in structured data settings, with clear implications for the feasibility and design of quantum vs. classical algorithms in ML tasks.
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 $A \in \mathbb{C}^{m\times n}$, a vector $b \in \mathbb{C}^{n}$, a bounded degree-$d$ polynomial $p$, and linear-time pre-processing, we can output a description of a vector $v$ such that $\|v - p(A) b\| \leq \varepsilon\|b\|$ in $\widetilde{\mathcal{O}}(d^{11} \|A\|_{\mathrm{F}}^4 / (\varepsilon^2 \|A\|^4 ))$ time. This improves upon the best known classical algorithm [CGLLTW, STOC'20, arXiv:1910.06151], which requires $\widetilde{\mathcal{O}}(d^{22} \|A\|_{\mathrm{F}}^6 /(\varepsilon^6 \|A\|^6 ) )$ 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 $p(A)b$ to $\varepsilon\|b\|$ error in $\widetilde{\mathcal{O}}(d\|A\|_{\mathrm{F}}/(\varepsilon \|A\|))$ 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.