Gaussian Process Regression: Combined kernels in different input spaces #19487
appletree999
started this conversation in
General
Replies: 1 comment 2 replies
-
I don't think our API supports this. You can probably write your own kernel classes to add this feature by subclassing the kernel classes in scikit-learn. |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi, is there a way in sklearn to specify which input dimensions a kernel works on for a Gaussian Process Regression?
Currently when you add two kernels in sklearn: k(x, x') = k1(x, x') + k2(x, x'), the kernels operate on the same input space, i.e. they operate on all the input dimensions altogether.
Is there a way to specify the input dimensions for each kernel? If you want k1 only works on the dimensions [0, 1] and k2 only works on the dimension [2], for example in GPy you can do:
k(x, x') = RBF(active_dims=[0,1]) + RBF(active_dims=[2])
Beta Was this translation helpful? Give feedback.
All reactions