We need documentation describing how to use the explainability features in ML.NET

[rogancarr]

We have some great explainability features in ML.NET, but they are hidden here and there. It would be great to have an explainability doc, similar to the ML.NET Cookbook, that showed how to use these features and what they do.

[JRAlexander]

This should be in docs. Lets work together on it.

[shmoradims]

PFI is covered here.

[shmoradims]

We also need to cover FCC and GAM as explained by Rogan here:

"Permutation Feature Importance (PFI), which gives a global picture of feature importance for a model, Feature Contribution Calculator (FCC), which gives a local picture of feature importance for a sample, and GAM, which are a class of interpretable models that give both a global and local explanation. Along with functioning code, we will write documentation to explain how the methods work and point to relevant literature, and create samples showing how to use the features and how to interpret the results."