Ambiguous marginal likelihood in a Non-stationary GP #13
Replies: 1 comment 2 replies
-
Hey @patel-zeel, After taking a quick look at the paper, I'm left slightly confused. In the last paragraph of Section 3 and below equation (5), the authors state that they set |
Beta Was this translation helpful? Give feedback.
-
I am trying to implement a Non-stationary variant of Gibbs kernel. I agree that the following equation (For the Global GP) seems correct as per the common marginal likelihood equations in GPs (at the end of page 6)
![image](https://user-images.githubusercontent.com/59758528/123537054-3fedcc00-d74b-11eb-9417-125e91c173b1.png)
But, why the marginal likelihood for local/latent GP (given below and on page 7 beginning) is not in the same format (Missing the first term compared to above)?
![image](https://user-images.githubusercontent.com/59758528/123537123-8511fe00-d74b-11eb-9450-d177b43515ad.png)
According to the above equation, this question is relevant because if I directly use
f_post(X_bar).logpdf(l_bar)
for latent GP, it would be an incorrect thing to do. I would greatly appreciate an explanation on this problem.Beta Was this translation helpful? Give feedback.
All reactions