Reinforcement Learning Workshop by Markus Liedl
Reinforcement Learning is a great idea: Let an automated agent discover a world on its own. The human specifies the goals, what it should reach, and the agent will learn by itself which sequence of actions accomplishes them.
The Reinforcement Learning Workshop is a one day course to get you started applying established RL algorithms:
- minimum of necessary theory
- policy vs. value function
- Q-learning
- temporal difference algorithms
- REINFORCE algorithms
- Deep Q-Learning (DQN) and some extensions
- A3C
Applications that we can work on:
- solving labyrinths (solutions can be nicely visualised)
- balancing a cart pole
- 2d blocks world (arranging blocks in some way)
- board games like backgammon and go
- driving a simulated vehicle
We'll start with base implementations of these applications and modify them for the sake of experimentation.
If you are interested in the training mail me at markus.liedl.training@gmail.com Normally the training will happen in or near Munich, Germany.
This workshop is available in English and German.
Related Deep Learning courses are