Month/Year of Graduation

5-2026

Degree Name

Bachelor of Science (B.S.)

Department

Computer Science

First Advisor

Dr. Alfredo Perez

Abstract

This capstone project investigates whether deterrence can emerge as a meaningful strategy within a zero-sum stochastic game using multi-agent reinforcement learning (MARL). After outlining core concepts in game theory and deterrence, the study models a simplified deterrence environment in which two minimax-Q agents repeatedly interact under uncertainty and adversarial incentives. The agents learn from rewards shaped by escalation costs, unilateral vulnerability, and the stabilizing benefits of restraint. Results show that both agents consistently converge toward a conservative, status-quo strategy, overwhelmingly selecting the Maintain action while avoiding both escalation and restraint in most scenarios. This behavior reflects the risk-averse logic of deterrence and demonstrates that deterrence-consistent patterns can arise naturally from worst-case learning dynamics. The project provides a proof of concept for simulating strategic stability with MARL and highlights directions for more realistic deterrence modeling.

Download

Share

COinS