Cooperative Game Theory in Machine Learning

• A. Mascareñas-Pazos ^[1] ; S. Lorenzo-Freire ^[1] ; J.M. Alonso-Meijide ^[2]
  1. [1] Universidade da Coruña

     Universidade da Coruña

     A Coruña, España

     
  2. [2] Universidade de Santiago de Compostela

     Universidade de Santiago de Compostela

     Santiago de Compostela, España

     
• Localización: Proceedings XoveTIC 2024: Impulsando el talento científico / coord. por Manuel Lagos Rodríguez, Tirso Varela Rodeiro, Javier Pereira-Loureiro , Manuel Francisco González Penedo , 2024, págs. 177-182
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • 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.