Learning Unitaries by Gradient Descent

Abstract

We study the hardness of learning unitary transformations in $U(d)$ via gradient descent on time parameters of alternating operator sequences. We provide numerical evidence that, despite the non-convex nature of the loss landscape, gradient descent always converges to the target unitary when the sequence contains $d^2$ or more parameters. Rates of convergence indicate a "computational phase transition." With less than $d^2$ parameters, gradient descent converges to a sub-optimal solution, whereas with more than $d^2$ parameters, gradient descent converges exponentially to an optimal solution.

References in corpus (10)

• A variational eigenvalue solver on a quantum processor
• A Quantum Approximate Optimization Algorithm
• Quantum-assisted quantum compiling
• Quantum Supremacy through the Quantum Approximate Optimization Algorithm
• A Quantum Approximate Optimization Algorithm Applied to a Bounded Occurrence Constraint Problem
• Quantum approximate optimization is computationally universal
• Searching for quantum optimal controls under severe constraints
• Searching for an optimal control in the presence of saddles on the quantum mechanical observable landscape
• Quantum Control Landscapes Are Almost Always Trap Free
• Efficient implementation of unitary transformations

Cited by in corpus (22)

• Challenges and Opportunities in Quantum Machine Learning
• Layerwise learning for quantum neural networks
• Diagnosing Barren Plateaus with Tools from Quantum Optimal Control
• Quantum agents in the Gym: a variational quantum algorithm for deep Q-learning
• Theory of overparametrization in quantum neural networks
• Differentiable Quantum Architecture Search
• Large gradients via correlation in random parameterized quantum circuits
• A comprehensive review of Quantum Machine Learning: from NISQ to Fault Tolerance
• Learning quantum data with the quantum Earth Mover's distance
• Universal Effectiveness of High-Depth Circuits in Variational Eigenproblems
• Progress toward favorable landscapes in quantum combinatorial optimization
• Algebraic Bethe Circuits
• An Expressive Ansatz for Low-Depth Quantum Approximate Optimisation
• Classification of dynamical Lie algebras for translation-invariant 2-local spin systems in one dimension
• Entanglement Diagnostics for Efficient Quantum Computation
• Characterization of variational quantum algorithms using free fermions
• Quantum Chaos and Circuit Parameter Optimization
• Experimental pairwise entanglement estimation for an N-qubit system :A machine learning approach for programming quantum hardware
• Energy Landscape Plummeting in Variational Quantum Eigensolver: Subspace Optimization, Non-iterative Corrections and Generator-informed Initialization for Improved Quantum Efficiency
• Policy Gradient Approach to Compilation of Variational Quantum Circuits
• Adiabatic quantum computing with parameterized quantum circuits
• Dynamical transition in controllable quantum neural networks with large depth