Analytic continuation via domain-knowledge free machine learning

Abstract

We present a machine-learning approach to a long-standing issue in quantum many-body physics, namely, analytic continuation. This notorious ill-conditioned problem of obtaining spectral function from imaginary time Green's function has been a focus of new method developments for past decades. Here we demonstrate the usefulness of modern machine-learning techniques including convolutional neural networks and the variants of stochastic gradient descent optimiser. Machine-learning continuation kernel is successfully realized without any 'domain-knowledge', which means that any physical 'prior' is not utilized in the kernel construction and the neural networks 'learn' the knowledge solely from 'training'. The outstanding performance is achieved for both insulating and metallic band structure. Our machine-learning-based approach not only provides the more accurate spectrum than the conventional methods in terms of peak positions and heights, but is also more robust against the noise which is the required key feature for any continuation technique to be successful. Furthermore, its computation speed is 10${}^4$-10${}^5$ times faster than maximum entropy method.

8 pages, 7 figures

References in corpus (13)

• Natural Language Processing (almost) from Scratch
• Quantum Monte Carlo Impurity Solver for Cluster DMFT and Electronic Structure Calculations in Adjustable Base
• Quantum-enhanced machine learning
• Hybridization expansion impurity solver: General formulation and application to Kondo lattice and two-orbital models
• Efficient and Accurate Machine-Learning Interpolation of Atomic Energies in Compositions with Many Species
• Experimental Quantum Hamiltonian Learning
• Accelerate Monte Carlo Simulations with Restricted Boltzmann Machines
• Machine learning for many-body physics: The case of the Anderson impurity model
• Pure density functional for strong correlations and the thermodynamic limit from machine learning
• Antiferromagnetism and the gap of a Mott insulator: Results from analytic continuation of the self-energy
• Charge self-consistent many-body corrections using optimized projected localized orbitals
• Analytical continuation of imaginary axis data using maximum entropy
• Maximum Quantum Entropy Method

Cited by in corpus (47)

• Machine learning and the physical sciences
• Electronic Density Response of Warm Dense Matter
• Artificial Neural Network Approach to the Analytic Continuation Problem
• Progress on stochastic analytic continuation of quantum Monte Carlo data
• Exploring QCD matter in extreme conditions with Machine Learning
• Reconstructing QCD Spectral Functions with Gaussian Processes
• Spectral Reconstruction with Deep Neural Networks
• Analytical Continuation of Matrix-Valued Functions: Carathéodory Formalism
• Finite-size effects in the reconstruction of dynamic properties from ab initio path integral Monte-Carlo simulations
• Sparse Modeling in Quantum Many-Body Problems
• Robust analytic continuation of Green's functions via projection, pole estimation, and semidefinite relaxation
• Machine learning-based inversion of nuclear responses
• Reconstructing spectral functions via automatic differentiation
• Rethinking the ill-posedness of the spectral function reconstruction -- why is it fundamentally hard and how Artificial Neural Networks can help
• Bosonic Nevanlinna Analytic Continuation
• Extending the average spectrum method: Grid points sampling and density averaging
• Non-perturbative strong coupling at timelike momenta
• SpM: Sparse modeling tool for analytic continuation of imaginary-time Green's function
• The QCD moat regime and its real-time properties
• Robust analytic continuation combining the advantages of the sparse modeling approach and Padé approximation
• Learned Optimizers for Analytic Continuation
• Neural network approach to reconstructing spectral functions and complex poles of confined particles
• Bryan's Maximum Entropy Method -- diagnosis of a flawed argument and its remedy
• Noise Enhanced Neural Networks for Analytic Continuation
• Machine learning spectral functions in lattice QCD
• Modeling the unphysical pseudomode model with physical ensembles: simulation, mitigation, and restructuring of non-Markovian quantum noise
• Renormalized Perturbation Theory for Fast Evaluation of Feynman Diagrams on the Real Frequency Axis
• Accelerated Continuous time quantum Monte Carlo method with Machine Learning
• Stochastic pole expansion method
• Estimates of the dynamic structure factor for the finite temperature electron liquid via analytic continuation of path integral Monte Carlo data
• Nevanlinna.jl: A Julia implementation of Nevanlinna analytic continuation
• Quasiparticle and fully self-consistent GW methods: an unbiased analysis using Gaussian orbitals
• Training biases in machine learning for the analytic continuation of quantum many-body Green's functions
• Connecting Tikhonov regularization to the maximum entropy method for the analytic continuation of quantum Monte Carlo data
• Neural network enhanced hybrid quantum many-body dynamical distributions
• TRIQS/Nevanlinna: Implementation of the Nevanlinna Analytic Continuation method for noise-free data
• Dual formulation of the maximum entropy method applied to analytic continuation of quantum Monte Carlo data
• Barycentric rational function approximation made simple: A fast analytic continuation method for Matsubara Green's functions
• Neural Network Analytic Continuation for Monte Carlo: Improvement by Statistical Errors
• Electron-boson spectral density functions of cuprates obtained from optical spectra via machine learning
• Quantum Monte Carlo simulation of BEC-impurity tunneling
• Autoencoder-based analytic continuation method for strongly correlated quantum systems
• Analytic Continuation of Noisy Data Using Adams Bashforth ResNet
• Improving the Deconvolution of Spectrum at Finite Temperature via Neural Network
• Application of radial basis functions neutral networks in spectral functions
• PyLIT: Reformulation and implementation of the analytic continuation problem using kernel representation methods
• The noiseless limit and improved-prior limit of the maximum entropy method and their implications for the analytic continuation problem