For with , the change of variables givesEquivalently,This is the scale-invariant prior as a measure. Because its integral over diverges, it is an improper prior, not a normalized probability distribution on that range.
Articles by others on the same topic
There are currently no matching articles.