Game Theory in Binary Classification: A Nash Equilibrium Decision Tree Model
Loading...
Date
2025
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ummto.faculté des sciences
Abstract
In this thesis, we present the method in [13] which combines game theory and machine learning to develop a new decision tree model for binary classification, called the Nash Equilibrium Decision Tree (NE-DT). Unlike standard decision trees, NE-DT treats the splitting process as a game between two players (left and right branches), where each tries to best separate the data. The Nash equilibrium is a key concept in game theory ,is used to find the optimal split, improving accuracy and robustness. The work explains basic game theory and machine learning concepts, then details the NE-DT algorithm with clear mathematical steps. A simple example using the Iris dataset shows how the model works in practice. The results prove that NE-DT effectively separates classes while being easy to understand.
Description
61f.;ill:30cm
Keywords
Game theory, Nash equilibrium, Decision trees, Binary classification, Machine learning