reward function

A reward function is a mathematical description of the incentives given to an agent in reinforcement learning. It quantifies the objective of the learning task, guiding the agent’s actions towards desirable outcomes by providing feedback on their performance.

14 papers