Model-Based RL

• Last lecture : learn Policy directly from experience
• Previous lectures : learn value function directly from experience
• This lecture : learn Model directly from experience
• and use planning to construct a value function or policy
• Integrate learning and planning into a single architecture

Model-Free RL

• No model
• Learn value function or policy from experience

Model-Based RL

• Learn a model from experience

• Plan value function from model

• value/policy를 통해 experience를 하고

• experience를 통해 model learning을 하며

• model을 통해 planning ( solve ) 를 해서 다시 value/policy를 수정한다.

Advantages of Model-Based RL