EFFECT OF MEMORY ON NASH EQUILIBRIUM BASED ON THEORY GAME USING MODELING-BASED AGENTS FOR THE EL FAROL BAR PROBLEM CASE STUDY

• Main Author: Al Solkajer Bahdar, Muhammad
• Format: Final Project
• Language: Indonesia
• Subjects: Fisika
• Online Access: https://digilib.itb.ac.id/gdl/view/46594
• Institution: Institut Teknologi Bandung

Physics is a study that collaborates theory and empiricism by using mathematics as its language. In essence, it aims is to provide predictions and model a phenomenon. With their high curiosity, physicists use physics to solve broader problems by collaborating with other disciplines. One fundamental and complex issue is regardimg human interaction. With the ability to model and predict human interactions, we can solve a variety of economic and social problems. One problem that can model human interactions well is The El Farol bar problem. This problem illustrates how humans as complexities involving others must choose a rational decision within their limited knowledge to get the most optimal benefits for themselves. Based on the game theory, everyone will race to come to the bar to enjoy Irish music if it's quiet and instead avoid coming to the bar if it's crowded based on their memory capacity and number of strategies. The system optimum solution will be obtained if the El Farol bar has a right level of crowd. The El Farol bar problem can be simulated using a repetitive agent-based modeling. The results of El Farol bar problem simulation provide a nash equilibrium. Reward analysis provides an overview of the behavior of each agent. Memory capacity provides equilibrium with low errors, while the number of strategies determines where the equilibrium is.