Certified Lower Bounds and Efficient Estimation of Minimum Accuracy in Quantum Kernel Methods

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gonçalves, Demerson N., Fernandes, Tharso D., Cordeiro, Andrias M. M., Lugao, Pedro H. G., Dias, João T.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914217722904576
author Gonçalves, Demerson N.
Fernandes, Tharso D.
Cordeiro, Andrias M. M.
Lugao, Pedro H. G.
Dias, João T.
author_facet Gonçalves, Demerson N.
Fernandes, Tharso D.
Cordeiro, Andrias M. M.
Lugao, Pedro H. G.
Dias, João T.
contents The minimum accuracy heuristic evaluates quantum feature maps without requiring full quantum support vector machine (QSVM) training. However, the original formulation is computationally expensive, restricted to balanced datasets, and lacks theoretical backing. This work generalizes the metric to arbitrary binary datasets and formally proves it constitutes a certified lower bound on the optimal empirical accuracy of any linear classifier in the same feature space. Furthermore, we introduce Monte Carlo strategies to efficiently estimate this bound using a random subset of Pauli directions, accompanied by rigorous probabilistic guarantees. These contributions establish minimum accuracy as a scalable, theoretically sound tool for pre-screening feature maps on near-term quantum devices.
format Preprint
id arxiv_https___arxiv_org_abs_2512_20588
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Certified Lower Bounds and Efficient Estimation of Minimum Accuracy in Quantum Kernel Methods
Gonçalves, Demerson N.
Fernandes, Tharso D.
Cordeiro, Andrias M. M.
Lugao, Pedro H. G.
Dias, João T.
Quantum Physics
Data Structures and Algorithms
The minimum accuracy heuristic evaluates quantum feature maps without requiring full quantum support vector machine (QSVM) training. However, the original formulation is computationally expensive, restricted to balanced datasets, and lacks theoretical backing. This work generalizes the metric to arbitrary binary datasets and formally proves it constitutes a certified lower bound on the optimal empirical accuracy of any linear classifier in the same feature space. Furthermore, we introduce Monte Carlo strategies to efficiently estimate this bound using a random subset of Pauli directions, accompanied by rigorous probabilistic guarantees. These contributions establish minimum accuracy as a scalable, theoretically sound tool for pre-screening feature maps on near-term quantum devices.
title Certified Lower Bounds and Efficient Estimation of Minimum Accuracy in Quantum Kernel Methods
topic Quantum Physics
Data Structures and Algorithms
url https://arxiv.org/abs/2512.20588