On Sep 12, 3:58 pm, "Michael Tobis" <[EMAIL PROTECTED]> wrote:
> I was amazed to see that this is a real controversy in some circles.
The bayesian vs. frequentist division is well alive, and will be until
the last frequentist is dead. :)
There are practicing frequentists who say that they are in principle
bayesian, but that because formulating one's prior in any more than
2-3 dimensions is impossible, bayesian methods are in practice
unusable. Pointing out the implicit priors behind frequentist methods
(confidence intervals, ridge regression), or suggesting that model
families are actually discrete priors does not influence their
opinion.
Meanwhile, bayesian methods are successfully used in machine learning
and in computational statistics with (kind of uninformative) priors
that are in practice confirmed empirically. And increasingly
hierarchical data forces bayesian thinking even on conservative fields
such as hypotheses testing in clinical research.
But even subjective probabilities are conditioned on some model. That
model needs to make sense and be explicit enough, if one wants the
probabilities to be taken seriously.
In the context of weather prediction etc., it is important to
understand that (1) bayesian techniques can be applied even when the
priors are not anyone's subjective beliefs; (2) bayesian models can be
tested and empirically validated just as any other models.
--
Janne
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
Global Change ("globalchange") newsgroup. Global Change is a public, moderated
venue for discussion of science, technology, economics and policy dimensions of
global environmental change.
Posts will be admitted to the list if and only if any moderator finds the
submission to be constructive and/or interesting, on topic, and not
gratuitously rude.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/globalchange
-~----------~----~----~----~------~----~------~--~---
