Adds BaselineImputer and Reworks MarginalImputer

[mmschlk]

TLDR: This PR closes #261 and this PR closes #107.

This pull request introduces several updates to the shapiq package, focusing on the addition of a new BaselineImputer class and various improvements to existing imputer classes.

New Features:

• Added BaselineImputer class for imputing missing values using predefined baseline values or calculated mean/mode from background data. (shapiq/games/imputer/baseline_imputer.py)

Updates to Existing Features:

• Removed the sample_replacements parameter from MarginalImputer and added a joint_marginal_distribution parameter. A DeprecationWarning for sample_replacements has been added. (shapiq/games/imputer/marginal_imputer.py)
• Enhanced the Imputer base class by adding new attributes (n_features, data, model, sample_size) and methods (fit, x property) to improve flexibility and functionality. (shapiq/games/imputer/base.py)

Codebase Organization:

• Updated imports and __all__ lists to include the new BaselineImputer class. (shapiq/__init__.py, shapiq/games/__init__.py, shapiq/games/imputer/__init__.py)

Documentation:

• Updated CHANGELOG.md to reflect the new features and changes, including the addition of BaselineImputer and the removal of the sample_replacements parameter from MarginalImputer.

Minor Improvements:

• Added a TODO comment in shapiq/games/base.py to consider removing a potentially runtime-increasing check. (shapiq/games/base.py)

[Advueu963]

Commented on some parameters, where the docs are missing.

Nice Code and great work!