notes.txt

## Reinforcement Learning Notes:

* Model-Based Learning (Dynamic Programming)
  * Value Iteration
  * Policy Iteration
  * https://s3-us-west-1.amazonaws.com/udacity-dlnfd/suttonbookdraft2018jan1.pdf
  * Chapters 4.1 - 4.3
* Model-Free Learning (Dynamic Programming)
  * Monte Carlo Methods
  * Temporal-Difference Learning
* RL in Continous Spaces
  * Deep Q-Learning (Value Based) / SARSA
    * These algorithms use maximization. Doesn't work in continuous spaces. 
  * Policy Gradients (Policy Based)
  * Actor-Critic Methods (Best of Both Worlds)
  * 2 ways to handle Continuous Spaces. Discretization and Function Approximation
* Watch David Silver's, from DeepMind, [reinforcement learning lecture series](https://www.youtube.com/watch?v=2pWv7GOvuf0&list=PLqYmG7hTraZDM-OYHWgPebj2MfCFzFObQ)
  * Lecture 7: Policy Gradient Methods
* Research the latest trends on Deep Reinforcement Learning
* http://pemami4911.github.io/blog/2016/08/21/ddpg-rl.html

## Asset files
* https://github.com/udacity/RoboND-QuadRotor-Unity-Simulator