Feature Synergy, Redundancy, and Independence in Global Model Explanations using SHAP Vector Decomposition — mentioned by BenoitP

Fri, 31 Dec 2021 08:45:09 GMT

Otherwise known as the Shapley value [1], which also takes into account the interaction between participants. > The Shapley value is the average marginal contribution of a feature value across all possible coalitions. [2] EDIT: There are efforts to decompose that attribution of value into Synergy, Redundancy and Independence components [3], which I believe could also have meaning the context of business. [1] https://en.wikipedia.org/wiki/Shapley_value [2] https://christophm.github.io/interpretable-ml-book/shapley.h... [3] https://arxiv.org/pdf/2107.12436.pdf

Paper and discussion history. A mention is not an endorsement.

Feature Synergy, Redundancy, and Independence in Global Model Explanations using SHAP Vector Decomposition · HackerPapers discussions

Feature Synergy, Redundancy, and Independence in Global Model Explanations using SHAP Vector Decomposition — mentioned by BenoitP