Implement heuristics for Shapley based on influence functions

As described in 4.5 of _Jia, Ruoxi, David Dao, Boxin Wang, Frances Ann Hubis, Nick Hynes, Nezihe Merve Gürel, Bo Li, Ce Zhang, Dawn Song, and Costas J. Spanos. ‘Towards Efficient Data Valuation Based on the Shapley Value’. In Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics, 1167–76. PMLR, 2019. http://proceedings.mlr.press/v89/jia19a.html._

The issue references a specific academic paper section (4.5) for implementing Shapley value heuristics using influence functions. Start by reading the cited paper to understand the proposed method. Then examine the existing pyDVL codebase for influence function and Shapley value implementations to see where the new heuristics should integrate. 'Done' means the heuristic is implemented, tested, and documented within the library's framework.

Rédigé par le modèle d'indexation à partir du texte de l'issue.