Chapter 1

1. Reinforcement learning (RL) is a branch of machine learning where the learning occurs via interacting with an environment.
2. RL works by train and error method, unlike other ML paradigms.
3. Agents are the software programs that make intelligent decisions and they are basically learners in RL.
4. Policy function specifies what action to take in each state and value function specifies the value of each state.
5. In model-based agent use the previous experience whereas in model-free learning there won't be any previous experience.
6. Deterministic, stochastic, fully observable, partially observable, discrete continuous, episodic and non-episodic.
7. OpenAI Universe provides rich environments for training RL agents.
8. Refer section Applications of RL.