Jeffreys prior

Named after the Bayesian statistician Harold Jeffreys.

The Jeffreys prior is a prior over the parameter space $\Theta$ is Bayesian statistics that is proportional to the determinant of the Fisher information matrix:

$$\pi (\theta )\propto \sqrt{|I(\theta )|}.$$

Jeffreys was motivated to find a prior which remained invariant under re-parameterizations of $\Theta$.