Automatic Tempered Posterior Distributions for Bayesian Inversion Problems

We propose a novel adaptive importance sampling scheme for Bayesian inversion problems where the inference of variables interest and power data noise are carried out using distinct (but interacting) methods. More specifically, we consider analysis (i.e., parameters model to invert), whereas employ maximum likelihood approach estimation power. The whole technique is implemented by means an iterative procedure with alternating optimization steps. Moreover, also used as tempered parameter posterior distribution interest. Therefore, sequence densities generated, automatically selected according current estimate A complete study over scale can be performed. Numerical experiments show benefits proposed approach.