Hi, When we call pm.find_MAP() on a marginal likelihood implementation of GPs (e.g. here ) I’m assuming, as per Bayesian Data Analysis, 3rd Ed. page 504 that the marginal likelihood is first maximized, then the (log) prior is added and then the maximum of the of posterior found? I’ve read however …

Kernel hyper-parameters priors and their use in MAP estimates of Gaussian Processes

tirthasheshpatel June 3, 2020, 1:10pm 4

Oh, now I follow. You are right! find_MAP does consider the conditional p(y \mid \theta, X) too and thus optimize for the values of hyper parameters \theta.

Topic		Replies	Views
What makes difference find_MAP() and pm.sample() Questions	2	2422	October 17, 2021
Gaussian Process Regression level 1 inference: Re-producing Mauna Loa CO2 Example with PyMC3 Questions	10	2244	August 16, 2017
Calculating marginal likelihood of GP Model Questions	6	1319	December 31, 2018
Unconsistent logp, Evaluating the results of find_MAP on Custom Likelihood Questions	3	610	August 28, 2018
Should find_MAP() be used for gaussian process models? Questions	1	622	May 22, 2020

Kernel hyper-parameters priors and their use in MAP estimates of Gaussian Processes

Related topics