Exercises
The following is a series of exercises. Answers to all the exercises are given in the Answers_to_Exercises_Chap14.ipynb Jupyter Notebook in the GitHub repository:
- The Matérn kernel function,
, can be thought of as a generalization of the RBF kernel. It is of the following form:
Eq. 29
is the modified Bessel function of the second kind. The Matérn kernel is specified by the parameters, 
and 
. The lengthscale parameter, , plays a similar role to the length-scale parameter, 
, in the RBF kernel in Eq. 10. The parameter, 
, controls how smooth the functions are when we use a GP prior with a Matérn covariance kernel.
Using the data from the code example in the main text and a Matérn kernel with the default value, 
, fit a GPR model to the data. Make predictions for a range of 
values. Note that for the Matérn kernel, the parameter, 
, is not optimized by the scikit-learn fitting process, so if you instantiate a Matérn kernel...
