Dynamic Agent-Based Reward Shaping for Multi-Agent Systems

Earlier works have reported that reward shaping accelerates the convergence of reinforcement learning algorithms. It also helps to make better use of existinginformation. In this article we propose the use to modify Qlearning in multi agent systems by the use of reward shaping depending on agent state regarding other agents. We studythis method with different choices, which indicate different effects of this method on the maze problem. The resultsindicate the directional search, reduces the number of steps to reach the target in the proposed modified approach if appropriate parameters are utilized

کلیدواژه ها:

multi-agent systems ، reward shaping ، agentbased learning