Second-order optimisation strategies for neural network quantum states

Abstract

The Variational Monte Carlo method has recently seen important advances through the use of neural network quantum states. While more and more sophisticated ansätze have been designed to tackle a wide variety of quantum many-body problems, modest progress has been made on the associated optimisation algorithms. In this work, we revisit the Kronecker-Factored Approximate Curvature, an optimiser that has been used extensively in a variety of simulations. We suggest improvements on the scaling and the direction of this optimiser, and find that they substantially increase its performance at a negligible additional cost. We also reformulate the Variational Monte Carlo approach in a game theory framework, to propose a novel optimiser based on decision geometry. We find that, on a practical test case for continuous systems, this new optimiser consistently outperforms any of the KFAC improvements in terms of stability, accuracy and speed of convergence. Beyond Variational Monte Carlo, the versatility of this approach suggests that decision geometry could provide a solid foundation for accelerating a broad class of machine learning algorithms.

35 pages, 9 figures, 4 tables. Accepted in Phil. Trans. R. Soc. A

References in corpus (11)

• A simple linear algebra identity to optimize Large-Scale Neural Network Quantum States
• Message-Passing Neural Quantum States for the Homogeneous Electron Gas
• A Kaczmarz-inspired approach to accelerate the optimization of neural network wavefunctions
• Neural Wave Functions for Superfluids
• Simulations of state-of-the-art fermionic neural network wave functions with diffusion Monte Carlo
• Second-order optimisation strategies for neural network quantum states
• Machine learning one-dimensional spinless trapped fermionic systems with neural-network quantum states
• Neural-network quantum states for ultra-cold Fermi gases
• Quantum natural gradient without monotonicity
• Fermi-Bose mapping and N-particle ground state of spin-polarized fermions in tight atom waveguides
• Distilling the essential elements of nuclear binding via neural-network quantum states

Cited by in corpus (5)

• Second-order optimisation strategies for neural network quantum states
• Efficient Optimization of Variational Autoregressive Networks with Natural Gradient
• Modal Backflow Neural Quantum States for Anharmonic Vibrational Calculations
• Digitized Counter-Diabatic Quantum Optimization for Bin Packing Problem
• Generalized Lanczos method for systematic optimization of neural-network quantum states