Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Bayesian model comparison #116

Open
stefanocovino opened this issue May 24, 2019 · 2 comments
Open

Bayesian model comparison #116

stefanocovino opened this issue May 24, 2019 · 2 comments

Comments

@stefanocovino
Copy link

Hi,

I am trying to derive some inference from regressions based on different kernel combinations, and I am doing so computing the Bayes factors (indeed by means of a Parallel-Tempering Ensemble MCMC to compute the evidence). This is fine however, as far as I have understood the matter, this requires to add priors (when they are needed) properly normalised. I just wonder whether the bounds one can introduce in the kernel definitions behave like flat priors or it is anyway better to explicitly define these priors and take care of their normalisations.

Thanks,
Stefano

@dfm
Copy link
Owner

dfm commented May 29, 2019

Good question! The priors returned by george are not properly normalized so you'd be better off doing it yourself.

@stefanocovino
Copy link
Author

Thanks! Good to know.

A kind suggestion. Adding to the docs an example with a properly written prior using george could anyway be useful, I think.

Bye,
Stefano

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants