Deep Reinforcement Learning for Epidemic Control of a Networked SIS Model

In this paper, we propose a new approach to reduce infection spread among individuals in a networked susceptible-infected-susceptible (SIS) model. Our approach assumes that each agent can adjust the strength of its connections with neighbors. An agent is randomly selected to learn about its local network and implement strategies to minimize infection spread over 100,000 episodes. We leverage deep deterministic policy gradient (DDPG), a deep reinforcement learning algorithm. We implement and evaluate our approach on graphs of various sizes and configurations. Additionally, we examine cooperative multi-agent scenarios in which agents collaborate to control the spread of infections.