Strategic Decision-Making in Oligopoly Markets Based on Stackelberg Game and Reinforcement Learning Algorithms

Advances in artificial intelligence have enabled the development of intelligent decision-making systems in dynamic and uncertain environments. This study proposes a novel framework that combines Stackelberg game theory with reinforcement learning to enhance the adaptability and optimization of hierarchical decision-making. The effectiveness of the framework is demonstrated through extensive experiments, showcasing the advantages of reinforcement learning algorithms in achieving faster convergence, better adaptability to dynamic conditions, and superior optimization performance. The Stackelberg game structure models the leader-follower interactions, allowing the leader firm to anticipate and influence the strategies of follower firms, while reinforcement learning empowers agents to adaptively optimize their strategies in response to dynamic market conditions. A simulation environment was designed to evaluate the proposed framework, where one leader and three followers competed to maximize their profits under conditions of fluctuating demand and external uncertainties. Experimental results demonstrate that the proposed approach significantly outperforms traditional methods, achieving higher profitability, faster convergence, and superior adaptability to market shocks. This research bridges the gap between theoretical game models and practical applications by embedding game-theoretic principles into machine learning frameworks, providing a robust solution for decision-making in uncertain and competitive markets. The findings have practical implications for intelligent economic systems and real-time optimization in similar domains. Future work can focus on extending the framework to multi-agent scenarios with heterogeneous participants and improving computational efficiency to scale the approach to more complex environments.