| |
BGonline.org Forums
A better approch?
Posted By: Maik Stiebler In Response To: A better approch? (Timothy Chow)
Date: Wednesday, 4 November 2009, at 10:51 p.m.
This is still a Bayesian viewpoint, so it does nothing to stop an "anti-frequentist rant."
I'm aware of that.
Also, I think you meant to say that the posterior distribution converges to a normal distribution, rather than is a normal distribution.
No, that's not what I meant.
once you start adding extra complications (e.g., variance reduction)
I assumed variance reduction is in place.
So I would still recommend computing posterior probabilities the way I have been indicating all along
You did not yet specify how to deal with VR.
since I suspect that it leads you to invest more computational effort
Why? You always need to integrate at some point to calculate the Bayesian answers to interesting questions. Maybe if you're doing what you described a few month ago, integrating is computationally cheaper because your posterior is discrete and somewhat poorly resolved.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.