Random Natural Gradient
arXiv:2311.04135 · doi:10.22331/q-2024-10-22-1503
Abstract
Hybrid quantum-classical algorithms appear to be the most promising approach for near-term quantum applications. An important bottleneck is the classical optimization loop, where the multiple local minima and the emergence of barren plateaux make these approaches less appealing. To improve the optimization the Quantum Natural Gradient (QNG) method [Quantum 4, 269 (2020)] was introduced - a method that uses information about the local geometry of the quantum state-space. While the QNG-based optimization is promising, in each step it requires more quantum resources, since to compute the QNG one requires quantum state preparations, where is the number of parameters in the parameterized circuit. In this work we propose two methods that reduce the resources/state preparations required for QNG, while keeping the advantages and performance of the QNG-based optimization. Specifically, we first introduce the Random Natural Gradient (RNG) that uses random measurements and the classical Fisher information matrix (as opposed to the quantum Fisher information used in QNG). The essential quantum resources reduce to linear and thus offer a quadratic "speed-up", while in our numerical simulations it matches QNG in terms of accuracy. We give some theoretical arguments for RNG and then benchmark the method with the QNG on both classical and quantum problems. Secondly, inspired by stochastic-coordinate methods, we propose a novel approximation to the QNG which we call Stochastic-Coordinate Quantum Natural Gradient that optimizes only a small (randomly sampled) fraction of the total parameters at each iteration. This method also performs equally well in our benchmarks, while it uses fewer resources than the QNG.
27 pages, 10 figures, v3 published version
References in corpus (32)
- Variational Quantum Algorithms
- Hardware-efficient Variational Quantum Eigensolver for Small Molecules and Quantum Magnets
- Ising formulations of many NP problems
- Noisy intermediate-scale quantum (NISQ) algorithms
- Predicting Many Properties of a Quantum System from Very Few Measurements
- Evaluating analytic gradients on quantum hardware
- Quantum Fisher information matrix and multiparameter estimation
- Variational ansatz-based quantum simulation of imaginary time evolution
- Quantum Natural Gradient
- The randomized measurement toolbox
- Quantum Approximate Optimization Algorithm for MaxCut: A Fermionic View
- Obstacles to State Preparation and Variational Optimization from Symmetry Protection
- The Quantum Approximate Optimization Algorithm and the Sherrington-Kirkpatrick Model at Infinite Size
- Beyond Barren Plateaus: Quantum Variational Algorithms Are Swamped With Traps
- Improving Variational Quantum Optimization using CVaR
- Optimal measurements for simultaneous quantum estimation of multiple phases
- Estimating the gradient and higher-order derivatives on quantum hardware
- Avoiding local minima in variational quantum eigensolvers with the natural gradient optimizer
- Performance of hybrid quantum/classical variational heuristics for combinatorial optimization
- Fisher Information in Noisy Intermediate-Scale Quantum Applications
- Quantum Optimization with a Novel Gibbs Objective Function and Ansatz Architecture Search
- Measuring Analytic Gradients of General Quantum Evolution with the Stochastic Parameter Shift Rule
- Fast quantum circuit cutting with randomized measurements
- Quantum Analytic Descent
- Quantum natural gradient generalised to noisy and non-unitary circuits
- Natural Evolutionary Strategies for Variational Quantum Computation
- Variational Quantum Time Evolution without the Quantum Geometric Tensor
- An evolving objective function for improved variational quantum optimisation
- Effects of noise on the overparametrization of quantum neural networks
- Generalization of Quantum Machine Learning Models Using Quantum Fisher Information Metric
- Natural evolution strategies and variational Monte Carlo
- Natural parameterized quantum circuit