Avoiding local minima in variational quantum eigensolvers with the natural gradient optimizer

Abstract

We compare the BFGS optimizer, ADAM and Natural Gradient Descent (NatGrad) in the context of Variational Quantum Eigensolvers (VQEs). We systematically analyze their performance on the QAOA ansatz for the Transverse Field Ising Model (TFIM) as well as on overparametrized circuits with the ability to break the symmetry of the Hamiltonian. The BFGS algorithm is frequently unable to find a global minimum for systems beyond about 20 spins and ADAM easily gets trapped in local minima. On the other hand, NatGrad shows stable performance on all considered system sizes, albeit at a significantly higher cost per epoch. In sharp contrast to most classical gradient based learning, the performance of all optimizers is found to decrease upon seemingly benign overparametrization of the ansatz class, with BFGS and ADAM failing more often and more severely than NatGrad. Additional tests for the Heisenberg XXZ model corroborate the accuracy problems of BFGS in high dimensions, but they reveal some shortcomings of NatGrad as well. Our results suggest that great care needs to be taken in the choice of gradient based optimizers and the parametrization for VQEs.

arXiv comment16 pages, 6 figures, 1 table

References in corpus (25)

• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• Barren plateaus in quantum neural network training landscapes
• 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
• An adaptive variational algorithm for exact molecular simulations on a quantum computer
• From the Quantum Approximate Optimization Algorithm to a Quantum Alternating Operator Ansatz
• Quantum Approximate Optimization Algorithm: Performance, Mechanism, and Implementation on Near-Term Devices
• Efficient variational quantum simulator incorporating active error minimisation
• Variational ansatz-based quantum simulation of imaginary time evolution
• Quantum Natural Gradient
• ProjectQ: An Open Source Software Framework for Quantum Computing
• Quantum Approximate Optimization Algorithm for MaxCut: A Fermionic View
• Hybrid Quantum-Classical Approach to Quantum Optimal Control
• Simulating quantum many-body dynamics on a current digital quantum computer
• Efficient Symmetry-Preserving State Preparation Circuits for the Variational Quantum Eigensolver Algorithm
• $XY$-mixers: analytical and numerical results for QAOA
• Sequential minimal optimization for quantum-classical hybrid algorithms
• Reachability Deficits in Quantum Approximate Optimization
• Efficient variational simulation of non-trivial quantum states
• Methodology for replacing indirect measurements with direct measurements
• On the Universality of the Quantum Approximate Optimization Algorithm
• Introduction to the Bethe ansatz II
• Geometry of learning neural quantum states

Cited by in corpus (76)

• The Variational Quantum Eigensolver: a review of methods and best practices
• Training variational quantum algorithms is NP-hard
• General parameter-shift rules for quantum gradients
• Exploring entanglement and optimization within the Hamiltonian Variational Ansatz
• Exploiting symmetry in variational quantum machine learning
• Theory of overparametrization in quantum neural networks
• Estimating the gradient and higher-order derivatives on quantum hardware
• SU(2) hadrons on a quantum computer
• Fisher Information in Noisy Intermediate-Scale Quantum Applications
• A Comparison of Various Classical Optimizers for a Variational Quantum Linear Solver
• Approximate amplitude encoding in shallow parameterized quantum circuits and its application to financial market indicator
• Near-Term Quantum Computing Techniques: Variational Quantum Algorithms, Error Mitigation, Circuit Compilation, Benchmarking and Classical Simulation
• Recent advances for quantum classifiers
• Variational Quantum Algorithms for Computational Fluid Dynamics
• ADAPT-VQE is insensitive to rough parameter landscapes and barren plateaus
• Real time evolution for ultracompact Hamiltonian eigenstates on quantum hardware
• Hamiltonian variational ansatz without barren plateaus
• Avoiding barren plateaus via transferability of smooth solutions in Hamiltonian Variational Ansatz
• Using models to improve optimizers for variational quantum algorithms
• Error mitigation for variational quantum algorithms through mid-circuit measurements
• Empirical performance bounds for quantum approximate optimization
• Measurement cost of metric-aware variational quantum algorithms
• Quantum Analytic Descent
• Quantum natural gradient generalised to noisy and non-unitary circuits
• Quantum Optimization for the Graph Coloring Problem with Space-Efficient Embedding
• Progress toward favorable landscapes in quantum combinatorial optimization
• ResQNets: A Residual Approach for Mitigating Barren Plateaus in Quantum Neural Networks
• Gradient Ascent Pulse Engineering with Feedback
• An entanglement perspective on the quantum approximate optimization algorithm
• Trainability Enhancement of Parameterized Quantum Circuits via Reduced-Domain Parameter Initialization
• Variational Quantum Evolution Equation Solver
• Optimizing quantum circuits with Riemannian gradient flow
• Stochastic Gradient Line Bayesian Optimization for Efficient Noise-Robust Optimization of Parameterized Quantum Circuits
• Alleviating Barren Plateaus in Parameterized Quantum Machine Learning Circuits: Investigating Advanced Parameter Initialization Strategies
• Efficient variational synthesis of quantum circuits with coherent multi-start optimization
• Effects of noise on the overparametrization of quantum neural networks
• Reduction of finite sampling noise in quantum neural networks
• Using Differential Evolution to avoid local minima in Variational Quantum Algorithms
• Mitigating algorithmic errors in quantum optimization through energy extrapolation
• Quantum algorithms for the generalized eigenvalue problem
• Variational quantum solver employing the PDS energy functional
• Improved variational quantum eigensolver via quasi-dynamical evolution
• Simulation of adiabatic quantum computing for molecular ground states
• Natural parameterized quantum circuit
• A stabilizer framework for Contextual Subspace VQE and the noncontextual projection ansatz
• Calibrating the Classical Hardness of the Quantum Approximate Optimization Algorithm
• A Parameter Setting Heuristic for the Quantum Alternating Operator Ansatz
• Optimizing Variational Quantum Algorithms with qBang: Efficiently Interweaving Metric and Momentum to Navigate Flat Energy Landscapes
• Numerical hardware-efficient variational quantum simulation of a soliton solution
• Measurement-induced entanglement phase transitions in variational quantum circuits
• Natural Gradient Optimization for Optical Quantum Circuits
• Experimental quantum natural gradient optimization in photonics
• Rayleigh-Gauss-Newton optimization with enhanced sampling for variational Monte Carlo
• Equivariant Variational Quantum Eigensolver to detect Phase Transitions through Energy Level Crossings
• Convergence and Quantum Advantage of Trotterized MERA for Strongly-Correlated Systems
• Quantum Natural Gradient optimizer on noisy platforms: QAOA as a case study
• Random Natural Gradient
• Efficient Quantum Circuits based on the Quantum Natural Gradient
• Energy Landscape Plummeting in Variational Quantum Eigensolver: Subspace Optimization, Non-iterative Corrections and Generator-informed Initialization for Improved Quantum Efficiency
• Calculation of the moscovium ground-state energy by quantum algorithms
• Parent Hamiltonian as a benchmark problem for variational quantum eigensolvers
• Optimized numerical gradient and Hessian estimation for variational quantum algorithms
• Pitfalls when tackling the exponential concentration of parameterized quantum models
• Exploring new variational quantum circuit ansatzes for solving $SU(2)$ matrix models
• Regularizing quantum loss landscapes by noise injection
• Robustness of optimized numerical estimation schemes for noisy variational quantum algorithms
• Quantum natural gradient with thermal-state initialization
• Global optimization in variational quantum algorithms via dynamic tunneling method
• Batched Line Search Strategy for Navigating through Barren Plateaus in Quantum Circuit Training
• Universal Euler-Cartan Circuits for Quantum Field Theories
• Circuit connectivity boosts by quantum-classical-quantum interfaces
• Improving the trainability of VQE on NISQ computers for solving portfolio optimization using convex interpolation
• Role of overparametrization in quantum approximate optimization
• Utility-Scale Quantum State Preparation: Classical Training using Pauli Path Simulation
• Adaptive time Compressed QITE (ACQ) and its geometrical interpretation
• Topological and geometric patterns in optimal bang-bang protocols for variational quantum algorithms: application to the $XXZ$ model on the square lattice