Date: Wed, 5 Sep 2007 09:58:51 -0700
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
Subject: Interpreting Diagnostic Bayesian VOI
Content-Type: text/plain; charset="iso-8859-1"
I am a relative newbie to Bayesian networks and statistics....
I am researching Value of Information (VOI) in Bayesian diagnostic
methods such as using entropy based methods. One of the things that I
am struggling with is how to describe the concept of VOI to a
statistician. Would r or r-squared be a good substitute? P-value? What
is the best analog for VOI in mainstream statistics?
A search on the web does not reveal anything useful. So all
suggestions are very welcome