Thesis
Improving deep reinforcement learning through better benchmarking and tools
- Abstract:
-
Reinforcement Learning (RL) has been a powerful approach for training machines to perform tasks such as game playing and, more recently, solving mathematics and coding puzzles. However, despite these successes, it has suffered from numerous issues. First, algorithms in reinforcement learning commonly rely on a number of implementation details which are crucial for performance, often used to adapt tools originally designed for supervised learning to the very different world of reinforcement le...
Expand abstract
Actions
Access Document
- Files:
-
-
(Preview, Dissemination version, pdf, 17.4MB, Terms of use)
-
Authors
Contributors
+ Whiteson, S
- Institution:
- University of Oxford
- Division:
- MPLS
- Department:
- Computer Science
- Role:
- Supervisor
+ Foerster, J
- Institution:
- University of Oxford
- Division:
- MPLS
- Department:
- Engineering Science
- Role:
- Supervisor
- ORCID:
- 0000-0001-9688-2498
+
Engineering and Physical Sciences Research Council
More from this funder
- Grant:
- 2416682
- Programme:
- EPSRC Centre for Doctoral Training in Autonomous Intelligent Machines and Systems
- DOI:
- Type of award:
- DPhil
- Level of award:
- Doctoral
- Awarding institution:
- University of Oxford
- Language:
-
English
- Keywords:
- Subjects:
- Deposit date:
-
2026-07-15
- ARK identifier:
Terms of use
- Copyright holder:
- Benjamin Ellis
- Copyright date:
- 2026
If you are the owner of this record, you can report an update to it here: Report update to this record