Characterizing the loss landscape of variational quantum circuits

Abstract

Machine learning techniques enhanced by noisy intermediate-scale quantum (NISQ) devices and especially variational quantum circuits (VQC) have recently attracted much interest and have already been benchmarked for certain problems. Inspired by classical deep learning, VQCs are trained by gradient descent methods which allow for efficient training over big parameter spaces. For NISQ sized circuits, such methods show good convergence. There are however still many open questions related to the convergence of the loss function and to the trainability of these circuits in situations of vanishing gradients. Furthermore, it is not clear how "good" the minima are in terms of generalization and stability against perturbations of the data and there is, therefore, a need for tools to quantitatively study the convergence of the VQCs. In this work, we introduce a way to compute the Hessian of the loss function of VQCs and show how to characterize the loss landscape with it. The eigenvalues of the Hessian give information on the local curvature and we discuss how this information can be interpreted and compared to classical neural networks. We benchmark our results on several examples, starting with a simple analytic toy model to provide some intuition about the behavior of the Hessian, then going to bigger circuits, and also train VQCs on data. Finally, we show how the Hessian can be used to adjust the learning rate for faster convergence during the training of variational circuits.

References in corpus (18)

• Quantum Computing in the NISQ era and beyond
• A variational eigenvalue solver on a quantum processor
• Barren plateaus in quantum neural network training landscapes
• The theory of variational hybrid quantum-classical algorithms
• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• Quantum Circuit Learning
• Quantum Chemistry in the Age of Quantum Computing
• Evaluating analytic gradients on quantum hardware
• A high-bias, low-variance introduction to Machine Learning for physicists
• Expressibility and entangling capability of parameterized quantum circuits for hybrid quantum-classical algorithms
• Data re-uploading for a universal quantum classifier
• Quantum Generative Adversarial Networks for Learning and Loading Random Distributions
• An initialization strategy for addressing barren plateaus in parametrized quantum circuits
• An Artificial Neuron Implemented on an Actual Quantum Processor
• Methodology for replacing indirect measurements with direct measurements
• Perspective: Energy Landscapes for Machine Learning
• Geometry of learning neural quantum states
• QuCumber: wavefunction reconstruction with neural networks

Cited by in corpus (32)

• Variational Quantum Algorithms
• Noisy intermediate-scale quantum (NISQ) algorithms
• The Variational Quantum Eigensolver: a review of methods and best practices
• The power of quantum neural networks
• Capacity and quantum geometry of parametrized quantum circuits
• Higher Order Derivatives of Quantum Neural Networks with Barren Plateaus
• Recent advances for quantum classifiers
• A comprehensive review of Quantum Machine Learning: from NISQ to Fault Tolerance
• Reinforcement Learning for Many-Body Ground-State Preparation Inspired by Counterdiabatic Driving
• Subtleties in the trainability of quantum machine learning models
• Neural Error Mitigation of Near-Term Quantum Simulations
• Realizing a quantum generative adversarial network using a programmable superconducting processor
• Absence of barren plateaus in finite local-depth circuits with long-range entanglement
• Variational Quantum Classifiers Through the Lens of the Hessian
• Variational Quantum Evolution Equation Solver
• Non-trivial symmetries in quantum landscapes and their resilience to quantum noise
• Variational quantum algorithm for unconstrained black box binary optimization: Application to feature selection
• Training robust and generalizable quantum models
• Hessian-based toolbox for reliable and interpretable machine learning in physics
• Quantum Energy Landscape and VQA Optimization
• Variational Quantum Algorithms for Semidefinite Programming
• Quantum computing through the lens of control: A tutorial introduction
• Optimizing Variational Quantum Algorithms with qBang: Efficiently Interweaving Metric and Momentum to Navigate Flat Energy Landscapes
• Expressivity of Variational Quantum Machine Learning on the Boolean Cube
• Beyond MP2 initialization for unitary coupled cluster quantum circuits
• Characterization of variational quantum algorithms using free fermions
• A hybrid quantum algorithm to detect conical intersections
• Density Matrix Emulation of Quantum Recurrent Neural Networks for Multivariate Time Series Prediction
• Distributed quantum machine learning via classical communication
• Universal noise-precision relations in variational quantum algorithms
• Adiabatic quantum computing with parameterized quantum circuits
• Detecting underdetermination in parameterized quantum circuits