Cooperative Game Theory in Machine Learning

Abstract

One of the key challenges in constructing a machine learning model is to select the most relevant features for optimal performance, as too many features can diminish model's effectiveness. This article explores the application of Cooperative Game Theory to facilitate such selection. Specifically, we utilize the Shapley value, a well-known solution in cooperative games. The machine learning model is represented as a cooperative game, where the Shapley value assesses the contribution of individual features to the model's overall performance.