Observing and Understanding Agent’s Characteristics With Environmental Changes for Learning Agents

As technology progress and reinforcement learning is implemented in many domains, from gaming to stock trading, it has become more important to understand agent characteristics. How it approaches the changes in the environment to understand its decision-making and make the system more reliable. Behavioral game theory broadens typical game theory by taking care of how players sense the reward, rewards that other players receive, limit strategic reasoning, the impact of situations, the effects of information and, how players use the information for further learning. In this paper, we have presented a card game that would enable us to enhance the analysis of learning agents’ characteristics with environmental changes. Environment change will also bring challenges regarding agent cooperation. Sometimes taking such a strategy not to share agents’ previous experiences could make itself more reliable to other agents. Our approach will provide more space or knowledge about how to improve the agent.