A robotics company is exploring different machine learning techniques to improve the decision-making capabilities of its autonomous robots. The team is particularly interested in reinforcement learning as a way to enable the robots to learn from their interactions with the environment and improve performance over time. To ensure they understand the underlying mechanics and applicability of this approach, they need a clear explanation of how reinforcement learning works and how it differs from other machine learning methods. How does reinforcement learning work?