Replies: 1 comment
-
@azinoma this depends a bit more on your use-case - e.g. what model class are you using? In regression with Gaussian likelihood, in principle you can do a low-rank update of the posterior (but this isn't implemented in GPflow). In variational models, you can set the rows/columns of q_mu and q_sqrt so that adding new points doesn't change the posterior, and then optimize only the variational parameters. Again, this would require some custom code at the moment as far as I am aware. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I am wondering what would be the best way to add an additional data point to the GP, without re-optimizing it.
Currently I am doing something like:
This works, but I am wondering if there is a better or more proper way to do it.
Thanks a lot for your help.
Beta Was this translation helpful? Give feedback.
All reactions