# Maximal data information priors

This approach was proposed by Arnold Zellner in [1] and for a parameter of interest ${\displaystyle \theta \;\;}$ the rule is given by
${\displaystyle \pi (\theta )=\exp\{\int \ p(x\mid \theta )\log \,p(x\mid \theta )\,dx\},}$
where ${\displaystyle p(x\mid \theta )}$ is the data density function.