Home | Search | Help  
Home Page Università di Genova

Seminar Details


Date 4-12-2007
Time 14:30
Room/Location DISI-Sala Conferenze 3 piano
Title Maximum Entropy Method for inverse problems
Speaker Dott. Jean-Michel Loubes
Affiliation Institut de Mathematiques, Toulouse, France
Link http://www.math.ups-tlse.fr/
Abstract When facing an ill-posed problem with unknown operator, few methods can be applied. In particular, most of the regularization methods can not be used since the knowledge of the eigenvalues of the operator is necessary. Hence Bayesian method can provide a consistent method to build estimates for the inverse problem. Here we focus on a Bayesian reconstruction of the inverse function considered as a discrete measure. This method can be applied to a large variety of fields and we will give as examples the cases of seismic tomography and survey sampling.
Back to Seminars