Learned Optimizers for Analytic Continuation

Abstract

Traditional maximum entropy and sparsity-based algorithms for analytic continuation often suffer from the ill-posed kernel matrix or demand tremendous computation time for parameter tuning. Here we propose a neural network method by convex optimization and replace the ill-posed inverse problem by a sequence of well-conditioned surrogate problems. After training, the learned optimizers are able to give a solution of high quality with low time cost and achieve higher parameter efficiency than heuristic fully-connected networks. The output can also be used as a neural default model to improve the maximum entropy for better performance. Our methods may be easily extended to other high-dimensional inverse problems via large-scale pretraining.

11 pages, 7 figures, 6 tables

References in corpus (9)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Continuous-time Monte Carlo methods for quantum impurity models
• Nevanlinna Analytical Continuation
• Sparse Modeling in Quantum Many-Body Problems
• Analytical continuation of imaginary axis data using maximum entropy
• Statistical and computational intelligence approach to analytic continuation in Quantum Monte Carlo
• The Average Spectrum Method for Analytic Continuation: Efficient Blocked Modes Sampling and Dependence on Discretization Grid
• Analytic continuation with Padé decomposition
• Effective classical correspondence of the Mott transition

Cited by in corpus (4)

• Progress on stochastic analytic continuation of quantum Monte Carlo data
• Noise Enhanced Neural Networks for Analytic Continuation
• Spectroscopic data de-noising via training-set-free deep learning method
• Training biases in machine learning for the analytic continuation of quantum many-body Green's functions