Reply from Bob O'Hara

Bob O'Hara — Thu, 21 Feb 2008 06:45:55 -0000

No replies yet? :-)

David Anderson has a new book out which, I’m sure, discusses information theory and likelihood. K-L divergence is the basis of a host of information criteria that we use to compare models.

Alternatively, you could take the Bayesian route, and have a look at Jaynes’ work, which is closely related to entropy methods.

Bob

Recent replies to "Log-likelihood ratio and information theory"

Reply from Bob O'Hara