Make classification GRUD more robust, and update docs

[WenjieDu]

What does this PR do?

1. fill in still-missing values of the empirical mean in GRUD with 0 to make it more robust with massive missing data;
2. update docs;

Before submitting

• This PR is made to fix a typo or improve the docs (you can dismiss the other checks if this is the case).
• Was this discussed/approved via a GitHub issue? Please add a link to it if that's the case.
• I have commented my code, particularly in hard-to-understand areas.
• I have written necessary tests and already run them locally.