I have created a hierarchical model that can be scaled in dimensions. Since there are multiple combinations of the right sets of parameters I tend to archive better results with NUTS sampler instead of SMC. (example below) For Model Selection I want to use the Bayes Factor. I know that the marginal…

Bayes Factor using NUTS sampler

junpenglao January 7, 2021, 8:54am 4

You can try bridge sampling (i have discussed with @Olaf in Fix Bridge sampling for variables with shape>1?).

1 Like

Topic		Replies	Views
Differenc between PyMC 3.1 and PyMC 3.0 on NUTS Questions	1	1285	August 28, 2017
Computation of Bayes Fractor Questions	3	698	April 20, 2020
Pm.step_methods() doesn't include pm.SMC() in pymc3 v 3.8 Questions	6	793	April 17, 2020
Slow speed of NUTS in Hierarchical model Questions	14	1978	October 20, 2020
How does pyMC handle the curse of dimensionality? version agnostic modeling	3	314	February 4, 2024