[Fis] Information, Bayesian

Dear FIS members:

I just found the following gem,
from Prof. Debabrata Basu:

****************************
Information is what it does for you:
It changes your opinion.
****************************

This is my best answer to Hans:

>(From Hans Christian von Baeyer):
>
>Julio, I think you would do us all a great service if you took your
>definition, which I agree benefits from its Bayeseian roots, and
>translatedit into simple words an/or symbols.
>
>In other words, can you give a definition that is in between Soren's
>oracular pronouncement, and Kullbach/Leibler math jargon?

This (also oracular, or aphoristic) statement
of Basu captures the essence of what
information means for a Bayesian.

The following glossary may help:

opinion => posterior probability distribution
change => divergence measure (like K-L)

>Maybe using a simple example? I'm always trying to popularize.

As for an example, I have a project,
lets say 2 in the next years :-),
to write an introductory book
on Bayesian statistics...
Meanwhile, see DeGroot,
Probability and Statistics (**2nd** ed.)

>Thanks!! Hans

You are very welcome, Julio!!

_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis
Received on Wed Sep 14 16:43:36 2005