Course details

8.3487

Deep Reinforcement Learning

S
WS 2020 Dr. Elia Bruni
4h/wk
8 ECTS
B.Sc modules:
CS-BWP-AI - Artificial Intelligence
CS-BWP-NI - Neuroinformatics
KOGW-WPM-KI - Artificial Intelligence
KOGW-WPM-NI - Neuroinformatics
M.Sc modules:
CC-MWP-NI - Neuroinformatics
CS-MWP-NI - Neuroinformatics
KOGW-MWPM-NIR - Major subject Neuroinformatics and Robotics

Doctorate program

The course gives an introduction to the field of (Deep) Reinforcement Learning. In four blocks it discusses (1) Basics or RL, including theory and basics of RL, based on the textbook "Reinforcement Learning: An Introduction" (Barto, Sutton), including topics like MDPs, SARSA, Q-Learning and Eligibility Traces. Block (2) introduces basic Deep RL approaches, DQN, DDPG and (V)PG. Block (3) tackles state of the art DRL algorithms, including TD3, SAC, TRPO, PPO, A2C, A3C, HER and off policy actor critic architectures. The courses concludes (4) with an outlook on advanced and specialized topics and a tutor supported final project. Students participating in the course will gain a broad overview over the field and be able to understand papers from the field and implement solutions making use of State of the Art Reinforcement Learning approaches. Preconditions: High School Level Math (High School / Introductory Math courses), Deep Learning(e.g IANNwTF, Machine Learning), ability to implement Deep Learning Algorithms in Python (at least Scientific Python, but actual experience in Tensorflow, e.g. IANNwTF is recommended). Previous experience with Reinforcement Learning (e.g. Methods of AI, Machine Learning) is useful but not necessary.