Help with Bayesian model averaging

Asked May 14 '23 at 06:00

Active May 14 '23 at 06:13

Viewed 43 times

I am new to statistical modelling and I need some help with Bayesian model averaging.
I have 3 models and I would like to derive a BMA of these models.

I am using the BIC estimate for the different models and a uniform prior probability of 0.33.

Model 1, BIC= 4420 Model 2, BIC= 3940 Model 3, BIC = 4325

I am stuck trying to calculate the Posterior probability for each model and I don’t know how to use that in estimating the BMA. Any help would be greatly appreciated.

edited May 14 '23 at 06:13

asked May 14 '23 at 06:00

Grace

You cannot derive the posterior probability from the BIC. – Xi'an May 14 '23 at 13:39
You could approximate $Pr(M_{i}|\mathbf{X})$ using the relative difference in BICs, as described in the answer to this related question. – Durden May 14 '23 at 23:31
@Durden, many thanks – Grace May 16 '23 at 10:19

Help with Bayesian model averaging

0 Answers0