A new influence measure for dependent features in classification problems
L. Davila-Pena, A. Saavedra Nieves, B. V. Casas Méndez
In this work, we introduce a novel approach to quantifying the influence of features on classification outcomes while accounting for dependencies among features. By adapting properties from cooperative game theory, we provide an axiomatic characterization and demonstrate that our measure generalizes the Banzhaf-Owen value for games with a priori unions. We also propose a sampling method for estimating the measure in large-scale datasets. Finally, the practical relevance of our proposals is illustrated through diverse applications.
Palabras clave: Classification problems, Influence measure of features, Dependent features, Axiomatic characterization, Banzhaf-Owen value
Programado
Jovenes Investigadores en teoría de juegos
12 de junio de 2025 11:30
MR 2
Otros trabajos en la misma sesión
A. Mascareñas, S. M. Lorenzo Freire, J. M. Alonso Meijide
I. Núñez Lugilde, I. M. García Jurado
A. Atay, C. Trudeau
J. C. Gonçalves Dosantos, R. Martínez, J. Sánchez-Soriano