Curator's Take
This article marks the first time a quantum‑enhanced physics‑informed neural network has been tailored to tackle non‑local problems such as integro‑differential and fractional PDEs, extending the reach of QML beyond ordinary differential equations. By proving an $L^{2}$ universal approximation bound for variational circuits with trigonometric feature maps and offering two concrete architectures—one that leverages high‑order quadrature and another that recasts the problem as a coupled PDE system—it shows how quantum hardware could directly encode the global operators that dominate many scientific models. If scalable implementations materialize, researchers may gain a new tool for simulating anomalous diffusion, viscoelasticity, or quantum transport with potentially fewer parameters than classical PINNs, although current results remain theoretical and await experimental validation on near‑term devices.
— Mark Eatherly
Summary
Quantum neural networks have emerged as powerful models for approximating nonlinear functions. Yet their use in solving integro-differential equations (IDEs) and fractional integro-partial differential equations (FIPDEs), which involve inherently nonlocal operators, remains unexplored. This work introduces a quantum physics-informed neural network (QPINN) framework that combines a quantum neural network with the governing equations of general nonlinear IDEs and FIPDEs. The proposed quantum network uses an affine feature map and variational quantum circuits to produce trial solutions with explicit trigonometric structure. We prove a quantitative $L^{2}(μ)$ universal approximation theorem for this architecture, achieving a convergence rate of $\mathcal{O}(n^{-1/2})$. This extends classical Fourier approximation theory to quantum circuits for physics-informed learning. We propose two QPINN variants: the numerical-quadrature QPINN (N-QPINN), which handles nonlocal integrals and fractional operators via high-order numerical quadrature while computing local derivatives through automatic differentiation of quantum trial solutions; and the auxiliary-function QPINN (A-QPINN), which eliminates numerical quadrature by introducing auxiliary variables that reformulate each integro-differential equation as an equivalent coupled system of partial differential equations, enabling a multi-output quantum neural network to simultaneously represent the solution and its associated variables. A series of numerical experiments demonstrates that the proposed QPINN framework accurately captures the behavior of nonlinear IDEs and FIPDEs and outperforms classical physics-informed neural networks.