Inference with minimal Gibbs free energy in information field theory

Abstract

Non-linear and non-Gaussian signal inference problems are difficult to tackle. Renormalization techniques permit us to construct good estimators for the posterior signal mean within information field theory (IFT), but the approximations and assumptions made are not very obvious. Here we introduce the simple concept of minimal Gibbs free energy to IFT, and show that previous renormalization results emerge naturally. They can be understood as being the Gaussian approximation to the full posterior probability, which has maximal cross information with it. We derive optimized estimators for three applications, to illustrate the usage of the framework: (i) reconstruction of a log-normal signal from Poissonian data with background counts and point spread function, as it is needed for gamma ray astronomy and for cosmography using photometric galaxy redshifts, (ii) inference of a Gaussian signal with unknown spectrum and (iii) inference of a Poissonian log-normal signal with unknown spectrum, the combination of (i) and (ii). Finally we explain how Gaussian knowledge states constructed by the minimal Gibbs free energy principle at different temperatures can be combined into a more accurate surrogate of the non-Gaussian posterior.

14 pages

References in corpus (4)

• Fast Hamiltonian sampling for large scale structure inference
• Bayesian power-spectrum inference for Large Scale Structure data
• Cosmic Cartography of the Large-Scale Structure with Sloan Digital Sky Survey Data Release 6
• Recovering the nonlinear density field from the galaxy distribution with a Poisson-Lognormal filter

Cited by in corpus (30)

• An improved map of the Galactic Faraday sky
• Estimating extragalactic Faraday rotation
• Reconstruction of Gaussian and log-normal fields with spectral smoothness
• NIFTY - Numerical Information Field Theory - a versatile Python library for signal inference
• IMAGINE: A comprehensive view of the interstellar medium, Galactic magnetic fields and cosmic rays
• Reconstruction of signals with unknown spectra in information field theory with parameter uncertainty
• RESOLVE: A new algorithm for aperture synthesis imaging of extended emission in radio astronomy
• Charting nearby dust clouds using Gaia data only
• The Galactic Faraday depth sky revisited
• Reply to "Comment on `Inference with minimal Gibbs free energy in information field theory'" by Iatsenko, Stefanovska and McClintock
• Information theory for fields
• D${}^3$PO - Denoising, Deconvolving, and Decomposing Photon Observations
• Faraday synthesis: The synergy of aperture and rotation measure synthesis
• Reconstructing signals from noisy data with unknown signal and noise covariance
• Probing magnetic helicity with synchrotron radiation and Faraday rotation
• A new approach to multi-frequency synthesis in radio interferometry
• Optimal Belief Approximation
• Information field dynamics for simulation scheme construction
• Generic inference of inflation models by non-Gaussianity and primordial power spectrum reconstruction
• Improving stochastic estimates with inference methods: calculating matrix diagonals
• Simulation of stochastic network dynamics via entropic matching
• Tomography of the Galactic free electron density with the Square Kilometer Array
• A Reputation Game Simulation: Emergent Social Phenomena from Information Theory
• Denoising, deconvolving and decomposing multi-domain photon observations- The D4PO algorithm
• Information Field Theory and Artificial Intelligence
• Dynamic system classifier
• Operator Calculus for Information Field Theory
• Improving self-calibration
• Quantifying imperfect cognition via achieved information gain
• HMCF - Hamiltonian Monte Carlo Sampling for Fields - A Python framework for HMC sampling with NIFTy