Add User Guide to the Docs

[mmschlk]

To welcome new users and summarize the field of Shapley-based explanations (and the contribution of shapiq), we should add a user guide to the documentation. The user guide should be written in an accessible but not overly colloquial and only for newcomers. The user guide should combine theory and implementation.

Implementation ways

The user guide of scikit learn or numpy is an amazing comparison:

• sklearn is a real book one can read from front to back. The user guide is written next to the core documentation
• numpy on the other hand places the user guide inside its API reference in very large docstrings.
  I don’t know what works best yet and fits shapiq best.

Notes

• The user guide should link to the correct modules and and methods
• the user guide should be like a book
• the user guide should use references where necessary and uphold scientific standards