The WhiRL Reading Group takes places weekly on Tuesdays, 15:00 at Tony Hoare room, Robert Hooke building.
Date of the next reading group: 29.01.2019
Topic: Deep Reinforcement Learning and the Deadly Triad
We know from reinforcement learning theory that temporal difference learning can fail in certain cases. Sutton and Barto (2018) identify a deadly triad of function approximation, bootstrapping, and off-policy learning. When these three properties are combined, learning can diverge with the value estimates becoming unbounded. However, several algorithms successfully combine these three properties, which indicates that there is at least a partial gap in our understanding. In this work, we investigate the impact of the deadly triad in practice, in the context of a family of popular deep reinforcement learning models – deep Q-networks trained with experience replay – analysing how the components of this system play a role in the emergence of the deadly triad, and in the agent’s performance.
Full text: Link
The reading group is open for all.
To add your email to the mailing list, please contact firstname.lastname@example.org