Least-informative Bayesian prior distributions for finite samples based on information theory
James C. Spall, Stacy D. Hill · IEEE Transactions on Automatic Control · 1990
A procedure is presented, based on Shannon information theory, for producing least-informative prior distributions for Bayesian estimation and identification. This approach relies on constructing an optimal mixture distribution and applies in small sample sizes (unlike certain approaches based on asymptotic theory). The procedure is illustrated in a small-scale numerical study and is contrasted with an approach based on maximum entropy.>