Thank you for your response. But, in the warning part, the answers are all true which mean that the computation of LOO/WAIC may not be reliable. Concerning the weights, how the best model doesn’t have the highest weight (probability) to be true?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Bayesian model averaging: ranking of model weights and LOO don't match | 10 | 1360 | March 14, 2020 | |
| Arviz compare: rank inconsistent with weight | 2 | 606 | June 24, 2022 | |
| Comparing models and selecting it with "weight" of arviz.compare | 1 | 1005 | June 10, 2021 | |
|
Comparing models - ranks vs weights
|
2 | 55 | July 30, 2024 | |
| Model comparison issue with LOO? | 14 | 2294 | January 30, 2018 |