Feature Synergy, Redundancy, and Independence in Global Model Explanations using SHAP Vector Decomposition
arXiv:2107.12436
Abstract
We offer a new formalism for global explanations of pairwise feature dependencies and interactions in supervised models. Building upon SHAP values and SHAP interaction values, our approach decomposes feature contributions into synergistic, redundant and independent components (S-R-I decomposition of SHAP vectors). We propose a geometric interpretation of the components and formally prove its basic properties. Finally, we demonstrate the utility of synergy, redundancy and independence by applying them to a constructed data set and model.
arXiv comment7 pages, 2 figures
References in corpus (6)
- Towards A Rigorous Science of Interpretable Machine Learning
- Consistent Individualized Feature Attribution for Tree Ensembles
- Explainable AI for Trees: From Local Explanations to Global Understanding
- Generating Counterfactual and Contrastive Explanations using SHAP
- True to the Model or True to the Data?
- Hodge decomposition and the Shapley value of a cooperative game