A Universal Training Algorithm for Quantum Deep Learning

Abstract

We introduce the Backwards Quantum Propagation of Phase errors (Baqprop) principle, a central theme upon which we construct multiple universal optimization heuristics for training both parametrized quantum circuits and classical deep neural networks on a quantum computer. Baqprop encodes error information in relative phases of a quantum wavefunction defined over the space of network parameters; it can be thought of as the unification of the phase kickback principle of quantum computation and of the backpropagation algorithm from classical deep learning. We propose two core heuristics which leverage Baqprop for quantum-enhanced optimization of network parameters: Quantum Dynamical Descent (QDD) and Momentum Measurement Gradient Descent (MoMGrad). QDD uses simulated quantum coherent dynamics for parameter optimization, allowing for quantum tunneling through the hypothesis space landscape. MoMGrad leverages Baqprop to estimate gradients and thereby perform gradient descent on the parameter landscape; it can be thought of as the quantum-classical analogue of QDD. In addition to these core optimization strategies, we propose various methods for parallelization, regularization, and meta-learning as augmentations to MoMGrad and QDD. We introduce several quantum-coherent adaptations of canonical classical feedforward neural networks, and study how Baqprop can be used to optimize such networks. We develop multiple applications of parametric circuit learning for quantum data, and show how to perform Baqprop in each case. One such application allows for the training of hybrid quantum-classical neural-circuit networks, via the seamless integration of Baqprop with classical backpropagation. Finally, for a representative subset of these proposed applications, we demonstrate the training of these networks via numerical simulations of implementations of QDD and MoMGrad.

arXiv comment83 pages, 45 figures

Cited by in corpus (30)

• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• Efficient Learning for Deep Quantum Neural Networks
• Quantum-assisted quantum compiling
• Quantum Approximate Optimization of the Long-Range Ising Model with a Trapped-Ion Quantum Simulator
• Solving nonlinear differential equations with differentiable quantum circuits
• Equivalence of quantum barren plateaus to cost concentration and narrow gorges
• Higher Order Derivatives of Quantum Neural Networks with Barren Plateaus
• Quantum autoencoders to denoise quantum data
• Quantum computing models for artificial neural networks
• Quantum Optimization with a Novel Gibbs Objective Function and Ansatz Architecture Search
• Reformulation of the No-Free-Lunch Theorem for Entangled Data Sets
• Information Scrambling in Quantum Neural Networks
• Quantum Graph Neural Networks
• Quantum Hamiltonian-Based Models and the Variational Quantum Thermalizer Algorithm
• Quantum Algorithms for Solving Ordinary Differential Equations via Classical Integration Methods
• Scrambling Ability of Quantum Neural Networks Architectures
• Non-trivial symmetries in quantum landscapes and their resilience to quantum noise
• A General Approach to Dropout in Quantum Neural Networks
• Entanglement-based quantum deep learning
• A comparative study of universal quantum computing models: towards a physical unification
• Avoiding local minima in Variational Quantum Algorithms with Neural Networks
• Data classification by quantum radial basis function networks
• A Continuous Variable Born Machine
• Quantum-enhanced neural networks in the neural tangent kernel framework
• The Hintons in your Neural Network: a Quantum Field Theory View of Deep Learning
• Permutation-equivariant quantum convolutional neural networks
• Experimental simulation of quantum superchannels
• Quantum Deformed Neural Networks
• A hybrid quantum-classical neural network with deep residual learning
• Learning to Utilize Correlated Auxiliary Noise: A Possible Quantum Advantage