Backpropagation scaling in parameterised quantum circuits

Abstract

The discovery of the backpropagation algorithm ranks among one of the most important moments in the history of machine learning, and has made possible the training of large-scale neural networks through its ability to compute gradients at roughly the same computational cost as model evaluation. Despite its importance, a similar backpropagation-like scaling for gradient evaluation of parameterised quantum circuits has remained elusive. Currently, the most popular method requires sampling from a number of circuits that scales with the number of circuit parameters, making training of large-scale quantum circuits prohibitively expensive in practice. Here we address this problem by introducing a class of structured circuits that are not known to be classically simulable and admit gradient estimation with significantly fewer circuits. In the simplest case -- for which the parameters feed into commuting quantum gates -- these circuits allow for fast estimation of the gradient, higher order partial derivatives and the Fisher information matrix. Moreover, specific families of parameterised circuits exist for which the scaling of gradient estimation is in line with classical backpropagation, and can thus be trained at scale. In a toy classification problem on 16 qubits, such circuits show competitive performance with other methods, while reducing the training cost by about two orders of magnitude.

Comments welcome

References in corpus (36)

• Variational Quantum Algorithms
• Barren plateaus in quantum neural network training landscapes
• Quantum Convolutional Neural Networks
• Parameterized quantum circuits as machine learning models
• Evaluating analytic gradients on quantum hardware
• The Variational Quantum Eigensolver: a review of methods and best practices
• Quantum metrology from a quantum information science perspective
• Quantum Natural Gradient
• An initialization strategy for addressing barren plateaus in parametrized quantum circuits
• Quantum convolutional neural network for classical data classification
• Layerwise learning for quantum neural networks
• Average-case complexity versus approximate simulation of commuting quantum computations
• Hybrid Quantum-Classical Approach to Quantum Optimal Control
• Measurement Optimization in the Variational Quantum Eigensolver Using a Minimum Clique Cover
• Efficient and Noise Resilient Measurements for Quantum Chemistry on Near-Term Quantum Computers
• Beyond Barren Plateaus: Quantum Variational Algorithms Are Swamped With Traps
• Diagnosing Barren Plateaus with Tools from Quantum Optimal Control
• Stochastic gradient descent for hybrid quantum-classical optimization
• Exploiting symmetry in variational quantum machine learning
• Efficient quantum measurement of Pauli operators in the presence of finite sampling error
• Large gradients via correlation in random parameterized quantum circuits
• A Lie Algebraic Theory of Barren Plateaus for Deep Parameterized Quantum Circuits
• Fisher Information in Noisy Intermediate-Scale Quantum Applications
• An Adaptive Optimizer for Measurement-Frugal Variational Algorithms
• Theory for Equivariant Quantum Neural Networks
• Theoretical Guarantees for Permutation-Equivariant Quantum Neural Networks
• The Adjoint Is All You Need: Characterizing Barren Plateaus in Quantum Ansätze
• Abrupt Transitions in Variational Quantum Circuit Training
• Progress toward favorable landscapes in quantum combinatorial optimization
• Classification of dynamical Lie algebras for translation-invariant 2-local spin systems in one dimension
• Escaping from the Barren Plateau via Gaussian Initializations in Deep Variational Quantum Circuits
• Lie-algebraic classical simulations for quantum computing
• Protocols for classically training quantum generative models on probability distributions
• Resource frugal optimizer for quantum machine learning
• Random coordinate descent: a simple alternative for optimizing parameterized quantum circuits
• Trade-off between Gradient Measurement Efficiency and Expressivity in Deep Quantum Neural Networks

Cited by in corpus (2)

• All you need is spin: SU(2) equivariant variational quantum circuits based on spin networks
• Learning complexity gradually in quantum machine learning models