Feature Dependence Challenges
Tim and Christoph delve into the complexities of feature dependence in model agnostic methods, highlighting how shared information can impact data manipulation and model predictions. They discuss the implications of breaking associations within data distributions and the potential pitfalls of using methods like Shapley values and lime in machine learning explanations.In this clip
From this podcast
Machine Learning Street Talk (MLST)
047 Interpretable Machine Learning - Christoph Molnar
Related Questions