Abstract:
This paper explores the integration of the Proximal Policy Optimization (PPO) algorithm in a 2D racing game created with Unity, aiming to demonstrate AI's transformative potential in gaming. By training an autonomous agent within the game’s environment, the study addresses challenges in environment design and algorithm optimization. It focuses on how PPO can enhance performance and adaptability in dynamic racing scenarios, offering insights into the synergy between game development and reinforcement learning.
Keywords: