CONSTRAINED NONINFORMATIVE PRIORS IN RISK ASSESSMENT

A constrained noninformative prior distribution, a generalization of t he Jeffreys noninformative prior, is defined for a single unknown para meter as the distribution corresponding to the maximum entropy distrib ution, subject to the assumed constraint(s), in the transformed model where the unknown parameter is approximately a location parameter. Thi s note illustrates this idea with binomial and Poisson data models, an d gives an example from risk assessment showing the practical usefulne ss of the constrained noninformative prior.