Data-driven Bayesian Deconvolution of Continuous Distributions of Relaxation Times

Le résumé fourni par la source

The knowledge of mechanical properties of materials is based on a precise analysis of their relaxation spectra. The development of methods to deconvolve spectra from measured data, and the assessment of their reliability, is therefore of paramount importance. We present a novel Bayesian deconvolution method based on a physically grounded parameterization of the spectra. We use a Metropolis-Hastings Markov-chain Monte Carlo fitting algorithm, with a full posterior analysis to obtain the best-fitting spectrum and its uncertainties. We test its performance on simulated data, finding that it is unbiased, reliable, and gives precise results even under strong noise.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.