Action
In reinforcement learning, the means by which an agent moves between different states in its environment. The agent selects which action to take using a policy. The action causes the agent to transition from its current state to a new state in the environment.
