Layerwise learning for quantum neural networks

Abstract

With the increased focus on quantum circuit learning for near-term applications on quantum devices, in conjunction with unique challenges presented by cost function landscapes of parametrized quantum circuits, strategies for effective training are becoming increasingly important. In order to ameliorate some of these challenges, we investigate a layerwise learning strategy for parametrized quantum circuits. The circuit depth is incrementally grown during optimization, and only subsets of parameters are updated in each training step. We show that when considering sampling noise, this strategy can help avoid the problem of barren plateaus of the error surface due to the low depth of circuits, low number of parameters trained in one step, and larger magnitude of gradients compared to training the full circuit. These properties make our algorithm preferable for execution on noisy intermediate-scale quantum devices. We demonstrate our approach on an image-classification task on handwritten digits, and show that layerwise learning attains an 8% lower generalization error on average in comparison to standard learning schemes for training quantum circuits of the same size. Additionally, the percentage of runs that reach lower test errors is up to 40% larger compared to training the full circuit, which is susceptible to creeping onto a plateau during training.

11 pages, 7 figures

References in corpus (8)

• Supplementary information for "Quantum supremacy using a programmable superconducting processor"
• A Quantum Approximate Optimization Algorithm
• Optimal Quantum Measurements of Expectation Values of Observables
• Learning to learn with quantum neural networks via classical neural networks
• Learning Unitaries by Gradient Descent
• Forward Thinking: Building and Training Neural Networks One Layer at a Time
• Accelerated variational algorithms for digital quantum simulation of many-body ground states
• A Jacobi Diagonalization and Anderson Acceleration Algorithm For Variational Quantum Algorithm Parameter Optimization

Cited by in corpus (150)

• Variational Quantum Algorithms
• The Variational Quantum Eigensolver: a review of methods and best practices
• The power of quantum neural networks
• Noise-Induced Barren Plateaus in Variational Quantum Algorithms
• Power of data in quantum machine learning
• Connecting ansatz expressibility to gradient magnitudes and barren plateaus
• A Review on Quantum Approximate Optimization Algorithm and its Variants
• Absence of Barren Plateaus in Quantum Convolutional Neural Networks
• General parameter-shift rules for quantum gradients
• Effect of barren plateaus on gradient-free optimization
• Diagnosing Barren Plateaus with Tools from Quantum Optimal Control
• Quantum Computing for High-Energy Physics: State of the Art and Challenges. Summary of the QC4HEP Working Group
• Quantum agents in the Gym: a variational quantum algorithm for deep Q-learning
• Solving nonlinear differential equations with differentiable quantum circuits
• Quantum circuit architecture search for variational quantum algorithms
• Equivalence of quantum barren plateaus to cost concentration and narrow gorges
• Higher Order Derivatives of Quantum Neural Networks with Barren Plateaus
• Filtering variational quantum algorithms for combinatorial optimization
• Quantum computing models for artificial neural networks
• Neutral Atom Quantum Computing Hardware: Performance and End-User Perspective
• Quantum Machine Learning: from physics to software engineering
• Theoretical Guarantees for Permutation-Equivariant Quantum Neural Networks
• Biology and medicine in the landscape of quantum advantages
• ADAPT-VQE is insensitive to rough parameter landscapes and barren plateaus
• Layer VQE: A Variational Approach for Combinatorial Optimization on Noisy Quantum Computers
• Barren plateaus in quantum tensor network optimization
• Quantum computing of the ${}^6$Li nucleus via ordered unitary coupled clusters
• Learning quantum data with the quantum Earth Mover's distance
• Reformulation of the No-Free-Lunch Theorem for Entangled Data Sets
• Avoiding barren plateaus via transferability of smooth solutions in Hamiltonian Variational Ansatz
• Generative quantum machine learning via denoising diffusion probabilistic models
• Realizing a quantum generative adversarial network using a programmable superconducting processor
• Reflection Equivariant Quantum Neural Networks for Enhanced Image Classification
• Universal Effectiveness of High-Depth Circuits in Variational Eigenproblems
• Training Saturation in Layerwise Quantum Approximate Optimisation
• Abrupt Transitions in Variational Quantum Circuit Training
• Avoiding Barren Plateaus with Classical Deep Neural Networks
• Hybrid Classical-Quantum Autoencoder for Anomaly Detection
• Hybrid quantum ResNet for car classification and its hyperparameter optimization
• Toward Trainability of Quantum Neural Networks
• Detecting and quantifying entanglement on near-term quantum devices
• Natural Evolutionary Strategies for Variational Quantum Computation
• Variational learning for quantum artificial neural networks
• ResQNets: A Residual Approach for Mitigating Barren Plateaus in Quantum Neural Networks
• Mitigating Barren Plateaus with Transfer-learning-inspired Parameter Initializations
• Applying QNLP to sentiment analysis in finance
• Comparative study of variations in quantum approximate optimization algorithms for the Traveling Salesman Problem
• Lyapunov control-inspired strategies for quantum combinatorial optimization
• Trainability Enhancement of Parameterized Quantum Circuits via Reduced-Domain Parameter Initialization
• Long-time simulations with high fidelity on quantum hardware
• A semi-agnostic ansatz with variable structure for quantum machine learning
• Uncovering Instabilities in Variational-Quantum Deep Q-Networks
• Entanglement entropy production in Quantum Neural Networks
• Variational Quantum Linear Solver with Dynamic Ansatz
• Classification of dynamical Lie algebras for translation-invariant 2-local spin systems in one dimension
• Hamiltonian operator approximation for energy measurement and ground state preparation
• Warm-Starting and Quantum Computing: A Systematic Mapping Study
• Engineered dissipation to mitigate barren plateaus
• Alleviating Barren Plateaus in Parameterized Quantum Machine Learning Circuits: Investigating Advanced Parameter Initialization Strategies
• On Circuit Depth Scaling For Quantum Approximate Optimization
• Learning shallow quantum circuits
• An evolving objective function for improved variational quantum optimisation
• Efficient variational synthesis of quantum circuits with coherent multi-start optimization
• Reduction of finite sampling noise in quantum neural networks
• The Quantum Path Kernel: a Generalized Quantum Neural Tangent Kernel for Deep Quantum Machine Learning
• Quantum computing quantum Monte Carlo with hybrid tensor network for electronic structure calculations
• Isometric tensor network optimization for extensive Hamiltonians is free of barren plateaus
• Here comes the SU(N): multivariate quantum gates and gradients
• Quantum Generative Learning for High-Resolution Medical Image Generation
• An exponentially-growing family of universal quantum circuits
• Networked Quantum Services
• Quantum Optimization for Training Quantum Neural Networks
• The quantum cost function concentration dependency on the parametrization expressivity
• Atomic Quantum Technologies for Quantum Matter and Fundamental Physics Applications
• The Unified Effect of Data Encoding, Ansatz Expressibility and Entanglement on the Trainability of HQNNs
• Improved variational quantum eigensolver via quasi-dynamical evolution
• Variational Preparation of the Sachdev-Ye-Kitaev Thermofield Double
• EHA: Entanglement-variational Hardware-efficient Ansatz for Eigensolvers
• Iterative Layerwise Training for Quantum Approximate Optimization Algorithm
• Long-lived Particles Anomaly Detection with Parametrized Quantum Circuits
• Quantum Computing and Tensor Networks for Laminate Design: A Novel Approach to Stacking Sequence Retrieval
• Evolution strategies: Application in hybrid quantum-classical neural networks
• A Parameter Setting Heuristic for the Quantum Alternating Operator Ansatz
• Numerical hardware-efficient variational quantum simulation of a soliton solution
• Randomized adaptive quantum state preparation
• Measurement-induced entanglement phase transitions in variational quantum circuits
• Variational Quantum Algorithms for Trace Distance and Fidelity Estimation
• Near-term Efficient Quantum Algorithms for Entanglement Analysis
• Progress towards analytically optimal angles in quantum approximate optimisation
• Forecasting steam mass flow in power plants using the parallel hybrid network
• Faster variational quantum algorithms with quantum kernel-based surrogate models
• Sequential optimal selection of a single-qubit gate and its relation to barren plateau in parameterized quantum circuits
• Generalization Error Bound for Quantum Machine Learning in NISQ Era -- A Survey
• Bayesian Learning of Parameterised Quantum Circuits
• Equivalence of cost concentration and gradient vanishing for quantum circuits: An elementary proof in the Riemannian formulation
• Protocols for Trainable and Differentiable Quantum Generative Modelling
• Barren plateaus are swamped with traps
• Random coordinate descent: a simple alternative for optimizing parameterized quantum circuits
• Performance of Particle Tracking Using a Quantum Graph Neural Network
• Quantum-enhanced neural networks in the neural tangent kernel framework
• Beyond Quantum Annealing: Optimal control solutions to MaxCut problems
• Information flow in parameterized quantum circuits
• Exponential Qubit Reduction in Optimization for Financial Transaction Settlement
• Mitigating Quantum Gate Errors for Variational Eigensolvers Using Hardware-Inspired Zero-Noise Extrapolation
• Machine-Learning Insights into the Entanglement-trainability Correlation of Parameterized Quantum Circuits
• Sketching the Best Approximate Quantum Compiling Problem
• QFCNN: Quantum Fourier Convolutional Neural Network
• Enhancing Variational Quantum Circuit Training: An Improved Neural Network Approach for Barren Plateau Mitigation
• Hamiltonian-based graph-state ansatz for variational quantum algorithms
• Single entanglement connection architecture between multi-layer bipartite Hardware Efficient Ansatz
• Backpropagation scaling in parameterised quantum circuits
• Restricting to the chip architecture maintains the quantum neural network accuracy
• Quantum Extremal Learning
• Leveraging commuting groups for an efficient variational Hamiltonian ansatz
• Learning to learn with an evolutionary strategy applied to variational quantum algorithms
• Almost fault-tolerant quantum machine learning with drastic overhead reduction
• An Amplitude-Based Implementation of the Unit Step Function on a Quantum Computer
• Perturbative gadgets for gate-based quantum computing: Non-recursive constructions without subspace restrictions
• Reducing measurement costs by recycling the Hessian in adaptive variational quantum algorithms
• Quantum Feasibility Labeling for NP-complete Vertex Coloring Problem
• VQE-generated quantum circuit dataset for machine learning
• Policy Gradient Approach to Compilation of Variational Quantum Circuits
• An Efficient Gradient Sensitive Alternate Framework for VQE with Variable Ansatz
• Barren plateaus are amplified by the dimension of qudits
• Quantum-secure multiparty deep learning
• A Universal Quantum Computer From Relativistic Motion
• Expressivity of deterministic quantum computation with one qubit
• Perspectives on Utilization of Measurements in Quantum Algorithms
• The Accuracy vs. Sampling Overhead Trade-off in Quantum Error Mitigation Using Monte Carlo-Based Channel Inversion
• Quantum Generative Training Using Rényi Divergences
• OnionVQE Optimization Strategy for Ground State Preparation on NISQ Devices
• Empirical Power of Quantum Encoding Methods for Binary Classification
• A Cyclic Layerwise QAOA Training
• A hybrid quantum-classical neural network with deep residual learning
• Mitigating Noise-Induced Gradient Vanishing in Variational Quantum Algorithm Training
• Generative flow-based warm start of the variational quantum eigensolver
• IGO-QNN: Quantum Neural Network Architecture for Inductive Grover Oracularization
• Improving Parameter Training for VQEs by Sequential Hamiltonian Assembly
• Orbital Expansion Variational Quantum Eigensolver: Enabling Efficient Simulation of Molecules with Shallow Quantum Circuit
• Batched Line Search Strategy for Navigating through Barren Plateaus in Quantum Circuit Training
• Adaptive Quantum Generative Training using an Unbounded Loss Function
• Performance Guarantees for Quantum Neural Estimation of Entropies
• FRQI Pairs method for image classification using Quantum Recurrent Neural Network
• Adaptive Interpolating Quantum Transform: A Quantum-Native Framework for Efficient Transform Learning
• Slice-Wise Initial State Optimization to Improve Cost and Accuracy of the VQE on Lattice Models
• Introducing Reduced-Width QNNs, an AI-inspired Ansatz Design Pattern
• Quantum Machine Learning for State Tomography Using Classical Data
• Gradient Scalability and Taylor Surrogation of Quantum Cost Landscapes
• Quantum feature-map learning with reduced resource overhead
• Compression of quantum shallow-circuit states