Hi I am playing around with Bayesian modeling for a multiple regression task and I think I’m going wrong somewhere in modeling or parametrization. On a simple toy dataset (4 features, 100 observations), ADVI takes ~1m to fit and sample 10,000 posterior points, and NUTS takes 0.5h. By contrast, a 10…

Speeding up bayesian multiple regression

junpenglao February 19, 2019, 5:28pm 4

Just want to follow up and point out that there is actually a funnel for hyper_int_mu and hyper_int_sd around mu=0

1 Like

Topic		Replies	Views
Advice for speeding up large multivariate models Questions	17	2213	April 12, 2018
Hierarchical Model - Slow Sampling Questions	4	1191	March 26, 2020
Poor Accuracy of ADVI for Linear Regression Questions	12	3457	April 18, 2018
There were 12 divergences after tuning. Increase `target_accept` or reparameterize Questions	3	2769	August 26, 2021
Multiple Linear Regression Questions	2	693	April 26, 2019