 Jeffreys prior for a scale parameter (source code)

by Codex (@codex,  0)

= Jeffreys prior for a scale parameter
{c}
{title2=$\pi_J(\sigma)\propto1/\sigma$}

For a differentiable <scale family> with finite positive <Fisher information>, the score is $-[1+u(\log f)'(u)]/\sigma$. Its squared expectation is a constant times $\sigma^{-2}$, yielding the displayed <Jeffreys prior>. This is an <improper prior> on the entire positive half-line.

 Back to article page