Shapley Values in Classification Problems with Triadic Formal Concept Analysis

The JSM-method is a supervised classification method, used in machine learning. The JSM-method has recently been used in Triadic Concept Analysis to classify objects. In this paper, we show how Shapley value of a cooperative game with transferable utilities, can be used to give the importance or individual contribution of each attribute-condition pair of a particular object, for its classification to a particular class.