Generalization in quantum machine learning from few training data

Abstract

Modern quantum machine learning (QML) methods involve variationally optimizing a parameterized quantum circuit on a training data set, and subsequently making predictions on a testing data set (i.e., generalizing). In this work, we provide a comprehensive study of generalization performance in QML after training on a limited number $N$ of training data points. We show that the generalization error of a quantum machine learning model with $T$ trainable gates scales at worst as $\sqrt{T/N}$. When only $K\ll T$ gates have undergone substantial change in the optimization process, we prove that the generalization error improves to $\sqrt{K/N}$. Our results imply that the compiling of unitaries into a polynomial number of native gates, a crucial application for the quantum computing industry that typically uses exponential-size training data, can be sped up significantly. We also show that classification of quantum states across a phase transition with a quantum convolutional neural network requires only a very small training data set. Other potential applications include learning quantum error correcting codes or quantum dynamical simulation. Our work injects new hope into the field of QML, as good generalization is guaranteed from few training data.

14+26 pages, 4+1 figures

References in corpus (27)

• An introduction to quantum machine learning
• Classical simulation of infinite-size quantum lattice systems in one spatial dimension
• Learning phase transitions by confusion
• Classical simulation of infinite-size quantum lattice systems in two spatial dimensions
• The quest for a Quantum Neural Network
• Hybrid quantum-classical algorithms and quantum error mitigation
• Resource-Aware Quantum Programming with General Recursion and Quantum Control
• Provably efficient machine learning for quantum many-body problems
• Concrete Categorical Model of a Quantum Circuit Description Language with Measurement
• Diagnosing Barren Plateaus with Tools from Quantum Optimal Control
• One-Dimensional Symmetry Protected Topological Phases and their Transitions
• Theory of overparametrization in quantum neural networks
• Generalization in Quantum Machine Learning: a Quantum Information Perspective
• Efficient measure for the expressivity of variational quantum algorithms
• Towards understanding the power of quantum kernels in the NISQ era
• Representation Learning via Quantum Neural Tangent Kernels
• Efficient classical simulation of the semi-classical Quantum Fourier Transform
• Robust Quantum Error Correction via Convex Optimization
• Structured Near-Optimal Channel-Adapted Quantum Error Correction
• Variational Hamiltonian Diagonalization for Dynamical Quantum Simulation
• Quantum Error Correction via Convex Optimization
• Variational Quantum Anomaly Detection: Unsupervised mapping of phase diagrams on a physical quantum computer
• Long-time simulations with high fidelity on quantum hardware
• Entangled Datasets for Quantum Machine Learning
• Experimental Quantum Learning of a Spectral Decomposition
• Unsupervised mapping of phase diagrams of 2D systems from infinite projected entangled-pair states via deep anomaly detection
• Rademacher complexity of noisy quantum circuits

Cited by in corpus (127)

• Challenges and Opportunities in Quantum Machine Learning
• Quantum advantage in learning from experiments
• Exploiting symmetry in variational quantum machine learning
• Theory for Equivariant Quantum Neural Networks
• Quantum Machine Learning: from physics to software engineering
• Theoretical Guarantees for Permutation-Equivariant Quantum Neural Networks
• Towards quantum enhanced adversarial robustness in machine learning
• Does provable absence of barren plateaus imply classical simulability?
• Understanding quantum machine learning also requires rethinking generalization
• Out-of-distribution generalization for learning quantum dynamics
• On the practical usefulness of the Hardware Efficient Ansatz
• Subtleties in the trainability of quantum machine learning models
• Benchmarking Adversarially Robust Quantum Machine Learning at Scale
• Quantum algorithms applied to satellite mission planning for Earth observation
• Machine Learning for Anomaly Detection in Particle Physics
• Analytic theory for the dynamics of wide quantum neural networks
• Generalization despite overfitting in quantum machine learning models
• Quantum phase detection generalisation from marginal quantum neural network models
• Benchmarking simulated and physical quantum processing units using quantum and hybrid algorithms
• Hybrid quantum ResNet for car classification and its hyperparameter optimization
• Quantum Machine Learning in Drug Discovery: Applications in Academia and Pharmaceutical Industries
• Quantum anomaly detection in the latent space of proton collision events at the LHC
• ResQNets: A Residual Approach for Mitigating Barren Plateaus in Quantum Neural Networks
• Pseudomagic Quantum States
• Tensor networks for quantum machine learning
• Applying QNLP to sentiment analysis in finance
• Overfitting in quantum machine learning and entangling dropout
• Towards quantum computing for clinical trial design and optimization: A perspective on new opportunities and challenges
• Dynamical simulation via quantum machine learning with provable generalization
• Quantum utility -- definition and assessment of a practical quantum advantage
• A semi-agnostic ansatz with variable structure for quantum machine learning
• Variational Quantum Approximate Support Vector Machine with Inference Transfer
• Learning quantum states and unitaries of bounded gate complexity
• Parallel Hybrid Networks: an interplay between quantum and classical neural networks
• Variational quantum one-class classifier
• How quantum computing can enhance biomarker discovery
• Learning Quantum Processes and Hamiltonians via the Pauli Transfer Matrix
• Reduction of finite sampling noise in quantum neural networks
• Generalization of Quantum Machine Learning Models Using Quantum Fisher Information Metric
• Resource Saving via Ensemble Techniques for Quantum Neural Networks
• Complexity of quantum circuits via sensitivity, magic, and coherence
• Numerical evidence against advantage with quantum fidelity kernels on classical data
• Quantum Convolutional Neural Networks are Effectively Classically Simulable
• Explaining Quantum Circuits with Shapley Values: Towards Explainable Quantum Machine Learning
• Dequantizing quantum machine learning models using tensor networks
• Enabling Large-Scale and High-Precision Fluid Simulations on Near-Term Quantum Computers
• A General Approach to Dropout in Quantum Neural Networks
• Transition Role of Entangled Data in Quantum Machine Learning
• Optimal quantum dataset for learning a unitary transformation
• Machine learning on quantum experimental data toward solving quantum many-body problems
• Approximately Equivariant Quantum Neural Network for $p4m$ Group Symmetries in Images
• What can we learn from quantum convolutional neural networks?
• Randomness-enhanced expressivity of quantum neural networks
• Quantum Vision Transformers for Quark-Gluon Classification
• Statistical Complexity of Quantum Learning
• Exploring Unsupervised Anomaly Detection with Quantum Boltzmann Machines in Fraud Detection
• QFactor: A Domain-Specific Optimizer for Quantum Circuit Instantiation
• A Quantum Kernel Learning Approach to Acoustic Modeling for Spoken Command Recognition
• Expressivity of Variational Quantum Machine Learning on the Boolean Cube
• When Federated Learning Meets Quantum Computing: Survey and Research Opportunities
• The role of data embedding in equivariant quantum convolutional neural networks
• Opportunities and challenges of quantum computing for climate modelling
• Hybrid quantum learning with data re-uploading on a small-scale superconducting quantum simulator
• Optimizing Quantum Convolutional Neural Network Architectures for Arbitrary Data Dimension
• Satellite image classification with neural quantum kernels
• Generalization Error Bound for Quantum Machine Learning in NISQ Era -- A Survey
• Quantum-machine-assisted Drug Discovery
• Exploring ab initio machine synthesis of quantum circuits
• Guided-SPSA: Simultaneous Perturbation Stochastic Approximation assisted by the Parameter Shift Rule
• Forecasting steam mass flow in power plants using the parallel hybrid network
• Error-tolerant quantum convolutional neural networks for symmetry-protected topological phases
• Guided Quantum Compression for High Dimensional Data Classification
• Application of quantum-inspired generative models to small molecular datasets
• The power and limitations of learning quantum dynamics incoherently
• Quantum Kernel Learning for Small Dataset Modeling in Semiconductor Fabrication: Application to Ohmic Contact
• Quantum Supervised Learning
• Variational Quantum Circuit Design for Quantum Reinforcement Learning on Continuous Environments
• Distributed quantum machine learning via classical communication
• Optimisation-free Classification and Density Estimation with Quantum Circuits
• Quantum neural compressive sensing for ghost imaging
• Robust Quantum Reservoir Computing for Molecular Property Prediction
• Alz-QNet: A Quantum Regression Network for Studying Alzheimer's Gene Interactions
• Error-mitigated photonic quantum circuit Born machine
• Quantum Kerr Learning
• High-fidelity dimer excitations using quantum hardware
• Ground state-based quantum feature maps
• Model-based Offline Quantum Reinforcement Learning
• Parallel-in-time quantum simulation via Page and Wootters quantum time
• Permutation-equivariant quantum convolutional neural networks
• Characterizing out-of-distribution generalization of neural networks: application to the disordered Su-Schrieffer-Heeger model
• Single-shot quantum machine learning
• Uncovering Quantum Many-body Scars with Quantum Machine Learning
• BCQQ: Batch-Constraint Quantum Q-Learning with Cyclic Data Re-uploading
• Supervised Learning Guarantee for Quantum AdaBoost
• QuForge: A Library for Qudits Simulation
• Deep Circuit Compression for Quantum Dynamics via Tensor Networks
• Quantum contextual bandits and recommender systems for quantum data
• Optimal lower bounds for Quantum Learning via Information Theory
• Exploiting subspace constraints and ab initio variational methods for quantum chemistry
• Learning Fourier series with parametrized quantum circuits
• Method for noise-induced regularization in quantum neural networks
• Application of Quantum Pre-Processing Filter for Binary Image Classification with Small Samples
• Transfer Learning Analysis of Variational Quantum Circuits
• Data-Dependent Generalization Bounds for Parameterized Quantum Models Under Noise
• Quantum Convolutional Neural Network for Phase Recognition in Two Dimensions
• Benchmarking Quantum Convolutional Neural Networks for Signal Classification in Simulated Gamma-Ray Burst Detection
• Out-of-distribution generalisation for learning quantum channels with low-energy coherent states
• Generalization Bounds in Hybrid Quantum-Classical Machine Learning Models
• Nearly query-optimal classical shadow estimation of unitary channels
• Conditional Generative Models for Learning Stochastic Processes
• High-Accuracy Temporal Prediction via Experimental Quantum Reservoir Computing in Correlated Spins
• Quantum Patch-Based Autoencoder for Anomaly Segmentation
• Guided Graph Compression for Quantum Graph Neural Networks
• Quantum Bayesian Networks for Machine Learning in Oil-Spill Detection
• Harnessing a 256-qubit Neutral Atom Simulator for Graph Classification
• Empirical Power of Quantum Encoding Methods for Binary Classification
• Integrated Encoding and Quantization to Enhance Quanvolutional Neural Networks
• Measuring Incompatible Observables with Quantum Neural Networks
• Quantum Phase Recognition via Quantum Attention Mechanism
• Quantum phase classification via partial tomography-based quantum hypothesis testing
• Disentangling quantum autoencoder
• Quantum Machine Learning for Climate Modelling
• Hybrid quantum recurrent neural network for remaining useful life prediction of turbofan engines
• Trainable Quantum Neural Network for Multiclass Image Classification with the Power of Pre-trained Tree Tensor Networks
• Transfer entropy and O-information to detect grokking in tensor network multi-class classification problems
• Learning complexity gradually in quantum machine learning models
• Vortex Detection from Quantum Data