The effect of data encoding on the expressive power of variational quantum machine learning models

Abstract

Quantum computers can be used for supervised learning by treating parametrised quantum circuits as models that map data inputs to predictions. While a lot of work has been done to investigate practical implications of this approach, many important theoretical properties of these models remain unknown. Here we investigate how the strategy with which data is encoded into the model influences the expressive power of parametrised quantum circuits as function approximators. We show that one can naturally write a quantum model as a partial Fourier series in the data, where the accessible frequencies are determined by the nature of the data encoding gates in the circuit. By repeating simple data encoding gates multiple times, quantum models can access increasingly rich frequency spectra. We show that there exist quantum models which can realise all possible sets of Fourier coefficients, and therefore, if the accessible frequency spectrum is asymptotically rich enough, such models are universal function approximators.

arXiv commentMinor corrections, including credits to a closely related paper. Source code available at https://github.com/XanaduAI/expressive_power_of_quantum_models

References in corpus (32)

• Supervised learning with quantum enhanced feature spaces
• Barren plateaus in quantum neural network training landscapes
• The theory of variational hybrid quantum-classical algorithms
• Quantum support vector machine for big data classification
• Quantum machine learning in feature Hilbert spaces
• Quantum Circuit Learning
• Quantum Convolutional Neural Networks
• Parameterized quantum circuits as machine learning models
• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• Circuit-centric quantum classifiers
• Expressibility and entangling capability of parameterized quantum circuits for hybrid quantum-classical algorithms
• Data re-uploading for a universal quantum classifier
• Quantum Natural Gradient
• Continuous-variable quantum neural networks
• Robust data encodings for quantum classifiers
• The Expressive Power of Parameterized Quantum Circuits
• Differentiable Learning of Quantum Circuit Born Machine
• PennyLane: Automatic differentiation of hybrid quantum-classical computations
• Structure optimization for parameterized quantum circuits
• Fantastic Generalization Measures and Where to Find Them
• Low-depth gradient measurements can improve convergence in variational hybrid quantum-classical algorithms
• Universal Variational Quantum Computation
• Universal discriminative quantum neural networks
• Quantum embeddings for machine learning
• Building quantum neural networks based on swap test
• Quantum Graph Neural Networks
• Quantum Hamiltonian-Based Models and the Variational Quantum Thermalizer Algorithm
• Pseudo-dimension of quantum circuits
• Summability of Multi-Dimensional Trigonometric Fourier Series
• Fourier Neural Networks: A Comparative Study
• Statistical Limits of Supervised Quantum Learning
• Quantum Circuit Parameters Learning with Gradient Descent Using Backpropagation

Cited by in corpus (255)

• Variational Quantum Algorithms
• The power of quantum neural networks
• Power of data in quantum machine learning
• Challenges and Opportunities in Quantum Machine Learning
• Generalization in quantum machine learning from few training data
• General parameter-shift rules for quantum gradients
• Is quantum advantage the right goal for quantum machine learning?
• Quantum machine learning beyond kernel methods
• Quantum agents in the Gym: a variational quantum algorithm for deep Q-learning
• Solving nonlinear differential equations with differentiable quantum circuits
• Exploiting symmetry in variational quantum machine learning
• Supervised quantum machine learning models are kernel methods
• Quantum machine learning for image classification
• Quantum Machine Learning for Chemistry and Physics
• Training Quantum Embedding Kernels on Near-Term Quantum Computers
• Quantum computing models for artificial neural networks
• Hybrid quantum neural network for drug response prediction
• The state of quantum computing applications in health and medicine
• Quantum Machine Learning: from physics to software engineering
• Recent advances for quantum classifiers
• Quantum Methods for Neural Networks and Application to Medical Image Classification
• Biology and medicine in the landscape of quantum advantages
• Universal Approximation Property of Quantum Machine Learning Models in Quantum-Enhanced Feature Spaces
• One qubit as a Universal Approximant
• Quantum Neural Network for Quantum Neural Computing
• Perceval: A Software Platform for Discrete Variable Photonic Quantum Computing
• Shadows of quantum machine learning
• Importance of Kernel Bandwidth in Quantum Machine Learning
• Machine Learning for Anomaly Detection in Particle Physics
• Enhancing Quantum Support Vector Machines through Variational Kernel Training
• Quantum machine learning of large datasets using randomized measurements
• Quantum Machine Learning for Digital Health? A Systematic Review
• Generalization despite overfitting in quantum machine learning models
• Classical surrogates for quantum learning models
• Quantum Analytic Descent
• Quantum Neural Network Classifiers: A Tutorial
• Generalized quantum circuit differentiation rules
• Variational inference with a quantum computer
• Quantum Kernel Methods for Solving Differential Equations
• Hybrid Classical-Quantum Autoencoder for Anomaly Detection
• Quantum Machine Learning for Finance
• Avoiding Barren Plateaus with Classical Deep Neural Networks
• Style-based quantum generative adversarial networks for Monte Carlo events
• Hybrid quantum ResNet for car classification and its hyperparameter optimization
• Quantum Kernel Methods under Scrutiny: A Benchmarking Study
• Quantum anomaly detection in the latent space of proton collision events at the LHC
• Efficient quantum amplitude encoding of polynomial functions
• High Dimensional Quantum Machine Learning With Small Quantum Computers
• Overfitting in quantum machine learning and entangling dropout
• Hybrid quantum physics-informed neural networks for simulating computational fluid dynamics in complex shapes
• Quantum Deep Hedging
• Tree tensor network classifiers for machine learning: from quantum-inspired to quantum-assisted
• Exponential data encoding for quantum supervised learning
• Quantum integration of elementary particle processes
• Generative machine learning with tensor networks: benchmarks on near-term quantum computers
• Quantum Deep Reinforcement Learning for Robot Navigation Tasks
• Fock State-enhanced Expressivity of Quantum Machine Learning Models
• Learning quantum states and unitaries of bounded gate complexity
• Variational quantum algorithm with information sharing
• Hybrid quantum cycle generative adversarial network for small molecule generation
• Multidimensional Fourier series with quantum circuits
• A brief introduction to quantum algorithms
• Uncovering Instabilities in Variational-Quantum Deep Q-Networks
• Biomarker Discovery with Quantum Neural Networks: A Case-study in CTLA4-Activation Pathways
• Quantum reservoir computing in finite dimensions
• On fundamental aspects of quantum extreme learning machines
• Entangled Datasets for Quantum Machine Learning
• Quantum Gaussian Process Regression for Bayesian Optimization
• Entanglement entropy production in Quantum Neural Networks
• Parallel Hybrid Networks: an interplay between quantum and classical neural networks
• Single-qubit universal classifier implemented on an ion-trap quantum device
• Fourier expansion in variational quantum algorithms
• A performance characterization of quantum generative models
• Hierarchical quantum circuit representations for neural architecture search
• Noisy Quantum Kernel Machines
• Optimal training of variational quantum algorithms without barren plateaus
• Quantum neural networks force fields generation
• Let Quantum Neural Networks Choose Their Own Frequencies
• Reduction of finite sampling noise in quantum neural networks
• Evaluation of Parameterized Quantum Circuits with Cross-Resonance Pulse-Driven Entanglers
• Tackling Sampling Noise in Physical Systems for Machine Learning Applications: Fundamental Limits and Eigentasks
• Evolutionary Quantum Architecture Search for Parametrized Quantum Circuits
• Quantum Convolutional Neural Networks with Interaction Layers for Classification of Classical Data
• Numerical evidence against advantage with quantum fidelity kernels on classical data
• A duplication-free quantum neural network for universal approximation
• Explaining Quantum Circuits with Shapley Values: Towards Explainable Quantum Machine Learning
• Dequantizing quantum machine learning models using tensor networks
• A General Approach to Dropout in Quantum Neural Networks
• Rapid training of quantum recurrent neural networks
• Data re-uploading with a single qudit
• Permutation Invariant Encodings for Quantum Machine Learning with Point Cloud Data
• Security Aspects of Quantum Machine Learning: Opportunities, Threats and Defenses
• HamLib: A library of Hamiltonians for benchmarking quantum algorithms and hardware
• Training robust and generalizable quantum models
• An exponentially-growing family of universal quantum circuits
• Neural Quantum Embedding: Pushing the Limits of Quantum Supervised Learning
• On the expressivity of embedding quantum kernels
• Robustness of quantum algorithms against coherent control errors
• Potential and limitations of random Fourier features for dequantizing quantum machine learning
• Predominant Aspects on Security for Quantum Machine Learning: Literature Review
• The Unified Effect of Data Encoding, Ansatz Expressibility and Entanglement on the Trainability of HQNNs
• The quantum cost function concentration dependency on the parametrization expressivity
• Quantum algorithms for scientific computing
• Ensemble-learning error mitigation for variational quantum shallow-circuit classifiers
• Conditional Born machine for Monte Carlo event generation
• VQC-Based Reinforcement Learning with Data Re-uploading: Performance and Trainability
• A Hyperparameter Study for Quantum Kernel Methods
• Trainability and Expressivity of Hamming-Weight Preserving Quantum Circuits for Machine Learning
• Statistical Complexity of Quantum Learning
• What can we learn from quantum convolutional neural networks?
• Randomness-enhanced expressivity of quantum neural networks
• Incremental Data-Uploading for Full-Quantum Classification
• Evolution strategies: Application in hybrid quantum-classical neural networks
• Simulation of a Variational Quantum Perceptron using Grover's Algorithm
• Neural quantum kernels: training quantum kernels with quantum neural networks
• Semisupervised Anomaly Detection using Support Vector Regression with Quantum Kernel
• Quantum kernels with squeezed-state encoding for machine learning
• Long-lived Particles Anomaly Detection with Parametrized Quantum Circuits
• Hyperparameter Importance of Quantum Neural Networks Across Small Datasets
• Variational Quanvolutional Neural Networks with enhanced image encoding
• Efficient MPS representations and quantum circuits from the Fourier modes of classical image data
• Quantum computing through the lens of control: A tutorial introduction
• Optimizing Quantum Convolutional Neural Network Architectures for Arbitrary Data Dimension
• Reinforcement learning-based architecture search for quantum machine learning
• Improving Gradient Methods via Coordinate Transformations: Applications to Quantum Machine Learning
• Opportunities and challenges of quantum computing for climate modelling
• Demonstration of a bosonic quantum classifier with data re-uploading
• Constrained and Vanishing Expressivity of Quantum Fourier Models
• When Federated Learning Meets Quantum Computing: Survey and Research Opportunities
• Expressivity of Variational Quantum Machine Learning on the Boolean Cube
• A Continuous Variable Born Machine
• Quantum Lazy Training
• Entanglement-induced provable and robust quantum learning advantages
• Pulsar Classification: Comparing Quantum Convolutional Neural Networks and Quantum Support Vector Machines
• Quantum Fourier Iterative Amplitude Estimation
• Measurable Krylov Spaces and Eigenenergy Count in Quantum State Dynamics
• Symmetry-invariant quantum machine learning force fields
• Generalization Error Bound for Quantum Machine Learning in NISQ Era -- A Survey
• Satellite image classification with neural quantum kernels
• Quantum Extreme Learning of molecular potential energy surfaces and force fields
• Forecasting steam mass flow in power plants using the parallel hybrid network
• Barren plateaus are swamped with traps
• Distributed Quantum Neural Networks via Partitioned Features Encoding
• Quantum data encoding as a distinct abstraction layer in the design of quantum circuits
• Guided Quantum Compression for High Dimensional Data Classification
• Evaluating the Resilience of Variational Quantum Algorithms to Leakage Noise
• Image Classification with Rotation-Invariant Variational Quantum Circuits
• Loop Feynman integration on a quantum computer
• Density Matrix Emulation of Quantum Recurrent Neural Networks for Multivariate Time Series Prediction
• Protocols for Trainable and Differentiable Quantum Generative Modelling
• Hybrid Quantum-Classical Autoencoders for End-to-End Radio Communication
• Variational Quantum Circuit Design for Quantum Reinforcement Learning on Continuous Environments
• The dilemma of quantum neural networks
• Performance of Particle Tracking Using a Quantum Graph Neural Network
• Exploration of Quantum Neural Architecture by Mixing Quantum Neuron Designs
• Photovoltaic power forecasting using quantum machine learning
• Expressive Quantum Supervised Machine Learning using Kerr-nonlinear Parametric Oscillators
• Hype or Heuristic? Quantum Reinforcement Learning for Join Order Optimisation
• Quantum Supervised Learning
• SnCQA: A hardware-efficient equivariant quantum convolutional circuit architecture
• Quantum-enhanced neural networks in the neural tangent kernel framework
• Gradients and frequency profiles of quantum re-uploading models
• Quantum Architecture Search for Quantum Monte Carlo Integration via Conditional Parameterized Circuits with Application to Finance
• A Quantum Algorithmic Approach to Multiconfigurational Valence Bond Theory: Insights from Interpretable Circuit Design
• On Optimizing Hyperparameters for Quantum Neural Networks
• Classical Ensembles of Single-Qubit Quantum Variational Circuits for Classification
• Non-Classical Kernels in Continuous Variable Systems
• The Inductive Bias of Quantum Kernels
• Quantum neural compressive sensing for ghost imaging
• Quantum density estimation with density matrices: Application to quantum anomaly detection
• Quantum Autoencoder for Multivariate Time Series Anomaly Detection
• Quantum Algorithm for Estimating Betti Numbers Using Cohomology Approach
• Quantum Machine Learning Playground
• Universal Approximation Theorem and error bounds for quantum neural networks and quantum reservoirs
• Quantum integration of decay rates at second order in perturbation theory
• On the similarity of bandwidth-tuned quantum kernels and classical kernels
• Ground state-based quantum feature maps
• Quantum Extremal Learning
• Deep tensor networks with matrix product operators
• Encoding-dependent generalization bounds for parametrized quantum circuits
• Restricting to the chip architecture maintains the quantum neural network accuracy
• BCQQ: Batch-Constraint Quantum Q-Learning with Cyclic Data Re-uploading
• Supervised Learning Guarantee for Quantum AdaBoost
• Model-based Offline Quantum Reinforcement Learning
• Quantum Adjoint Convolutional Layers for Effective Data Representation
• Data-Dependent Generalization Bounds for Parameterized Quantum Models Under Noise
• Quantum-inspired clustering with light
• Trained quantum neural networks are Gaussian processes
• QKAN: quantum Kolmogorov-Arnold networks with applications in machine learning and multivariate state preparation
• Optimized numerical gradient and Hessian estimation for variational quantum algorithms
• Quantum Neural Networks in Practice: A Comparative Study with Classical Models from Standard Data Sets to Industrial Images
• Approximation and Generalization Capacities of Parametrized Quantum Circuits for Functions in Sobolev Spaces
• Learning Fourier series with parametrized quantum circuits
• Noise tailoring for Robust Amplitude Estimation
• Fourier Analysis of Variational Quantum Circuits for Supervised Learning
• Multidimensional Quantum Generative Modeling by Quantum Hartley Transform
• Typical Machine Learning Datasets as Low-Depth Quantum Circuits
• A Survey of Classical And Quantum Sequence Models
• Crosstalk-Based Parameterized Quantum Circuit Approximation
• Quantum-Classical Autoencoder Architectures for End-to-End Radio Communication
• Quantum Active Learning
• Predicting Open-Hole Laminates Failure Using Support Vector Machines With Classical and Quantum Kernels
• Engineering Quantum Reservoirs through Krylov Complexity, Expressivity and Observability
• Expressivity of deterministic quantum computation with one qubit
• Quantitative convergence of trained quantum neural networks to a Gaussian process
• Generalization Bounds in Hybrid Quantum-Classical Machine Learning Models
• Policy Gradients using Variational Quantum Circuits
• A Hybrid Quantum-Classical Particle-in-Cell Method for Plasma Simulations
• Towards Improved Quantum Machine Learning for Molecular Force Fields
• Regularizing quantum loss landscapes by noise injection
• Quantum Physics-Informed Neural Networks for Maxwell's Equations: Circuit Design, "Black Hole" Barren Plateaus Mitigation, and GPU Acceleration
• Hybrid Quantum Classical Graph Neural Networks for Particle Track Reconstruction
• Expressiveness of Commutative Quantum Circuits: A Probabilistic Approach
• Quantum neural network autoencoder and classifier applied to an industrial case study
• An Empirical Study of Quantum Dynamics as a Ground State Problem with Neural Quantum States
• Estimation of mutual information via quantum kernel method
• Quantum approximated cloning-assisted density matrix exponentiation
• Benchmarking Quantum Surrogate Models on Scarce and Noisy Data
• Improving Parameter Training for VQEs by Sequential Hamiltonian Assembly
• When Quantum and Classical Models Disagree: Learning Beyond Minimum Norm Least Square
• Artificial intelligence for representing and characterizing quantum systems
• Multi-channel convolutional neural quantum embedding
• Universal approximation of continuous functions with minimal quantum circuits
• Empirical Power of Quantum Encoding Methods for Binary Classification
• Shot-based quantum encoding: a data-loading paradigm for quantum neural networks
• Interferometric binary phase estimations
• Quantum Sparse Coding and Decoding Based on Quantum Network
• Building Continuous Quantum-Classical Bayesian Neural Networks for a Classical Clinical Dataset
• Scaling of symmetry-restricted quantum circuits
• Double descent in quantum kernel methods
• Single-Qudit Quantum Neural Networks for Multiclass Classification
• Quantum Chebyshev Probabilistic Models for Fragmentation Functions
• New perspectives on quantum kernels through the lens of entangled tensor kernels
• Quantum feature-map learning with reduced resource overhead
• Exploiting biased noise in variational quantum models
• Pseudo quantum advantages in perceptron storage capacity
• Quantum-Enhanced Neural Exchange-Correlation Functionals
• Towards Continuous-variable Quantum Neural Networks for Biomedical Imaging
• Schmidt quantum compressor
• Error and Resource Estimates of Variational Quantum Algorithms for Solving Differential Equations Based on Runge-Kutta Methods
• Hybrid quantum recurrent neural network for remaining useful life prediction of turbofan engines
• Auto Quantum Machine Learning for Multisource Classification
• Detecting underdetermination in parameterized quantum circuits
• Data-driven Quantum Dynamical Embedding Method for Long-term Prediction on Near-term Quantum Computers
• Quantifying the Expressive Capacity of Quantum Systems: Fundamental Limits and Eigentasks
• Mean-field limit from general mixtures of experts to quantum neural networks
• Introducing Reduced-Width QNNs, an AI-inspired Ansatz Design Pattern
• QMC-Net: Data-Aware Quantum Representations for Remote Sensing Image Classification
• High-expressibility Quantum Neural Networks using only classical resources
• Extended parameter shift rules with minimal derivative variance for parameterized quantum circuits
• Quantum Machine Learning for Climate Modelling
• Efficient classical computation of the neural tangent kernel of quantum neural networks
• Learning complexity gradually in quantum machine learning models
• Quantum reservoir computing for predicting and characterizing chaotic maps
• Quantum Algorithms for Unsupervised Machine Learning and Neural Networks