Python Deep Learning - Second Edition

Types of RL agents

We have different types of RL agents:

• Value-based agents: These store the value function and base their decisions on it. Such an agent will decide which action to take based on the value of the states, where the action leads. These agents don't use a policy.
• Policy-based agents: These use only the policy, and not the value function, when deciding what action to take.
• Actor-critic agents: Use both the value function and the policy to make decisions.
• Model-based agents: These include a model of the environment. Given a state and an action, the agent can use the model as a simulation of the real environment to predict the next state and reward. In other words, the model allows the agent to plan its future actions.
• Model ...