A Game-Theoretic Analysis of the Recurrent Patient-Doctor Interaction Using Genetic Algorithms

Abstract

Medical service relies on patient-doctor interaction, which is modeled in this work with the Stag Hunt game. As prescribed by game theory, the Stag Hunt game has two pure-strategy Nash equilibria in which either both players cooperate or both players defect. This paper investigates means of achieving cooperation in a recurrent patient-doctor interaction scenario. For this purpose, genetic algorithms (GA) are employed to solve the Stag Hunt game. As such, evolutionism models learning along the generations, whereas randomness in the evolutionary process models trembles in the choice of strategy. Four test scenarios are simulated: variable number of GA iterations, variable initial choice of strategy, variable number of recurrent interactions and variable fraction of population partitioning. Simulation results provide valuable insights into the means of achieving patient-doctor cooperation.