brms implementation follows what I recommend, too. It has the additional benefit of using non-centered parameterization, which reduces the probability of getting funnel shaped posterior which is more difficult to sample.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| How to extent (regularized) horseshoe prior to multiple outcome setting? | 28 | 3821 | May 8, 2020 | |
| Horseshoe prior for logistic regression in brms | 0 | 1043 | July 12, 2018 | |
| Bayes Sparse Regression - reg. horseshoe for multinomial model | 17 | 4654 | August 28, 2019 | |
| Horseshoe regularization on matrix factorization not working | 4 | 234 | July 19, 2024 | |
| Adding expert recommended scores in regularized horseshoe priors | 1 | 466 | January 31, 2023 |