Product kernel explainer

[Advueu963]

Motivation and Context

This pull request tackles #425 and introduces a new model-specific explanation method, the ProductKernelExplainer, for product kernel-based machine learning models (such as Gaussian Processes and Support Vector Machines). The implementation includes new classes, conversion utilities, and kernel game logic, enabling exact computation of Shapley values for these models. Several supporting files and types are added, along with minor cleanups elsewhere in the codebase.

The main work was done by @IsaH57.

New Product Kernel Explainer functionality:

• Added ProductKernelExplainer class for model-specific explanations of product kernel-based models, supporting exact Shapley value computation for regression and classification models. Currently, it is restricted to methods using the RBF kernel.
• Implemented ProductKernelComputer class, including efficient Shapley value computation algorithms and kernel vector utilities.
• Added ProductKernelModel dataclass to standardise storage of kernel model parameters.
• Introduced conversion utilities to transform scikit-learn SVM and Gaussian Process models into the internal product kernel format.
• Implemented ProductKernelGame for evaluating the kernel-based cooperative game underlying the Shapley value computations.
• Added validation and type definitions for product kernel explainers and indices.

Public API Changes

• No Public API changes
• Yes, Public API changes (Details below)

We introduce a new Explainer, namely the ProductKernelExplainer, giving efficient computation of Shapley Values for kernel-based machine learning models.

How Has This Been Tested?

Unit and Integration Tests have been added to the test suite to ensure the correctness of the method.

Checklist

• The changes have been tested locally.
• Documentation has been updated (if the public API or usage changes).
• An entry has been added to CHANGELOG.md (if relevant for users).
• The code follows the project's style guidelines.
• I have considered the impact of these changes on the public API.

[codecov]

Codecov Report

❌ Patch coverage is 87.67773% with 26 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/shapiq/explainer/product_kernel/conversion.py	71.42%	10 Missing ⚠️
src/shapiq/explainer/product_kernel/explainer.py	85.71%	5 Missing ⚠️
src/shapiq/explainer/product_kernel/game.py	86.48%	5 Missing ⚠️
src/shapiq/explainer/product_kernel/validation.py	82.60%	4 Missing ⚠️
src/shapiq/explainer/product_kernel/base.py	92.85%	1 Missing ⚠️
.../shapiq/explainer/product_kernel/product_kernel.py	98.18%	1 Missing ⚠️

📢 Thoughts on this report? Let us know!

[mmschlk]

Amazing work! Love the tests, love the code, it's simple and understandable! I had three nitpicks only regarding the model typings, since I think this might be easily approved.