Provably efficient experimental classification of quantum data with quantum kernels
Abstract
Analyzing quantum systems is essential for quantum information processing, but can require exponentially many measurements without exploiting the structure of the system or task. Quantum machine learning offers a promising way to reduce this cost by learning from quantum data. Here, we theoretically establish a framework for efficient quantum-state classification and experimentally demonstrate the resulting scheme on a programmable photonic processor. Our analysis links learning efficiency to the structure of quantum data, without relying on a physical description tailored to each task. We derive sufficient conditions for accurate classification with a number of measurements polynomial in system size. We apply this guarantee to a resource problem in continuous-variable (CV) quantum computing: classifying stellar rank, a measure of non-Gaussianity. Experimentally, we implement the classification of single-mode CV quantum states with stellar rank zero and one. This is enabled by developing a programmable CV photonic processor capable of preparing and processing diverse Gaussian and non-Gaussian states and performing virtual projection onto such states, going beyond the Gaussian-only capabilities of most existing photonic platforms. The experiment demonstrates accurate classification despite experimental imperfections. Our results establish a theory-guided and experimentally validated framework for quantum-state classification that can be implemented on quantum devices available today.
Related
- Empirical Evaluation of Quantum Kernel Methods for Classification: A Comparative Study with Classical Kernels
- The Multi-Electron Unified Wave Equation in Quantum Chemistry: Symmetrized Riesz Exchange Kernels, Sobolev Domain Classification, and O(N4) Quasiparticle Electronic Structure across Bond Dissociation
- Provably Efficient Learning of Fermionic Correlations under Particle-Number Symmetry
- PQ-Vault: Provably Quantum-Safe Asset Custody for Unmodified Ethereum
- Quantum observables as Fréchet sensitivity kernels
BibTeX
@misc{anai2026provably,
title = {{Provably efficient experimental classification of quantum data with quantum kernels}},
author = {Keitaro Anai and Akimoto Nakayama and Shintaro Ooka and Hiroko Tomoda and Yuto Morohoshi and Yohji Chin and Shigehito Miki and Masahiro Yabuno and Hirotaka Terai and Peter van Loock and Kosuke Mitarai and Shuntaro Takeda},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.39350},
eprint = {2610.39350},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.39350},
}
RIS
TY - UNPB
TI - Provably efficient experimental classification of quantum data with quantum kernels
AU - Keitaro Anai
AU - Akimoto Nakayama
AU - Shintaro Ooka
AU - Hiroko Tomoda
AU - Yuto Morohoshi
AU - Yohji Chin
AU - Shigehito Miki
AU - Masahiro Yabuno
AU - Hirotaka Terai
AU - Peter van Loock
AU - Kosuke Mitarai
AU - Shuntaro Takeda
PY - 2026
DA - 2026/10/29
PB - arXiv
DO - 10.48550/arxiv.2610.39350
UR - https://arxiv.org/abs/2610.39350
AB - Analyzing quantum systems is essential for quantum information processing, but can require exponentially many measurements without exploiting the structure of the system or task. Quantum machine learning offers a promising way to reduce this cost by learning from quantum data. Here, we theoretically establish a framework for efficient quantum-state classification and experimentally demonstrate the resulting scheme on a programmable photonic processor. Our analysis links learning efficiency to the structure of quantum data, without relying on a physical description tailored to each task. We derive sufficient conditions for accurate classification with a number of measurements polynomial in system size. We apply this guarantee to a resource problem in continuous-variable (CV) quantum computing: classifying stellar rank, a measure of non-Gaussianity. Experimentally, we implement the classification of single-mode CV quantum states with stellar rank zero and one. This is enabled by developing a programmable CV photonic processor capable of preparing and processing diverse Gaussian and non-Gaussian states and performing virtual projection onto such states, going beyond the Gaussian-only capabilities of most existing photonic platforms. The experiment demonstrates accurate classification despite experimental imperfections. Our results establish a theory-guided and experimentally validated framework for quantum-state classification that can be implemented on quantum devices available today.
ER -