Implement custom kernel in gPyTorch


I am working on a project where I am using Gaussian Process regression to estimate the transition and measuerment functions of a system to be used in a Kalman Filter.

In order to do this I need to use the Jacobian of the learned models. This comes down to taking the derivative of the GP mean functions (and potentially covariance function). However, I’m not sure how to implement a GP model with custom kernel using gPyTorch.

Please let me know if I need to clarify anything.