Coalitional strategies for efficient individual prediction explanation
This work addresses the need for efficient and practical explanation methods to increase trust in ML models for end-users and stakeholders, though it is incremental as it builds on existing approaches like SHAP.
The paper tackles the problem of slow computation times and restrictive assumptions in existing methods for explaining individual predictions from machine learning models by introducing coalitional methods that detect relevant groups of attributes. The results show that these methods are more efficient than SHAP, shortening computation time while maintaining acceptable accuracy.
As Machine Learning (ML) is now widely applied in many domains, in both research and industry, an understanding of what is happening inside the black box is becoming a growing demand, especially by non-experts of these models. Several approaches had thus been developed to provide clear insights of a model prediction for a particular observation but at the cost of long computation time or restrictive hypothesis that does not fully take into account interaction between attributes. This paper provides methods based on the detection of relevant groups of attributes -- named coalitions -- influencing a prediction and compares them with the literature. Our results show that these coalitional methods are more efficient than existing ones such as SHapley Additive exPlanation (SHAP). Computation time is shortened while preserving an acceptable accuracy of individual prediction explanations. Therefore, this enables wider practical use of explanation methods to increase trust between developed ML models, end-users, and whoever impacted by any decision where these models played a role.