English

Quantum statistical learning via Quantum Wasserstein natural gradient

Mathematical Physics 2021-02-03 v1 Information Theory math.IT math.MP Optimization and Control Quantum Physics

Abstract

In this article, we introduce a new approach towards the statistical learning problem argminρ(θ)PθWQ2(ρ,ρ(θ))\operatorname{argmin}_{\rho(\theta) \in \mathcal P_{\theta}} W_{Q}^2 (\rho_{\star},\rho(\theta)) to approximate a target quantum state ρ\rho_{\star} by a set of parametrized quantum states ρ(θ)\rho(\theta) in a quantum L2L^2-Wasserstein metric. We solve this estimation problem by considering Wasserstein natural gradient flows for density operators on finite-dimensional CC^* algebras. For continuous parametric models of density operators, we pull back the quantum Wasserstein metric such that the parameter space becomes a Riemannian manifold with quantum Wasserstein information matrix. Using a quantum analogue of the Benamou-Brenier formula, we derive a natural gradient flow on the parameter space. We also discuss certain continuous-variable quantum states by studying the transport of the associated Wigner probability distributions.

Cite

@article{arxiv.2008.11135,
  title  = {Quantum statistical learning via Quantum Wasserstein natural gradient},
  author = {Simon Becker and Wuchen Li},
  journal= {arXiv preprint arXiv:2008.11135},
  year   = {2021}
}
R2 v1 2026-06-23T18:05:47.235Z