UTCS Artificial Intelligence
courses
talks/events
demos
people
projects
publications
software/data
labs
areas
admin
Batch Reinforcement Learning in a Complex Domain (2007)
Shivaram Kalyanakrishnan
and
Peter Stone
Temporal difference reinforcement learning algorithms are perfectly suited to autonomous agents because they learn directly from an agent's experience based on sequential actions in the environment. However, their most common algorithmic variants are relatively inefficient in their use of experience data, which in many agent-based settings can be scarce. In particular, they make just one learning ``update'' for each atomic experience. Batch reinforcement learning algorithms, on the other hand, aim to achieve greater data efficiency by saving experience data and using it in aggregate to make updates to the learned policy. Their success has been demonstrated in the past on simple domains like grid worlds and low-dimensional control applications like pole balancing. In this paper, we compare and contrast batch reinforcement learning algorithms with on-line algorithms based on their empirical performance in a complex, continuous, noisy, multiagent domain, namely RoboCup soccer Keepaway. We find that the two batch methods we consider, Experience Replay and Fitted Q Iteration, both yield significant gains in sample complexity, while achieving high asymptotic performance.
View:
PDF
,
PS
,
HTML
Citation:
In
The Sixth International Joint Conference on Autonomous Agents and Multiagent Systems
, pp. 650-657, New York, NY, USA, May 2007. ACM.
Bibtex:
@InProceedings{AAMAS07-kalyanakrishnan, title={Batch Reinforcement Learning in a Complex Domain}, author={Shivaram Kalyanakrishnan and Peter Stone}, booktitle={The Sixth International Joint Conference on Autonomous Agents and Multiagent Systems}, month={May}, address={New York, NY, USA}, publisher={ACM}, pages={650-657}, url="http://www.cs.utexas.edu/users/ai-lab?AAMAS07-kalyanakrishnan", year={2007} }
People
Shivaram Kalyanakrishnan
Ph.D. Alumni
shivaram [at] cs utexas edu
Peter Stone
Faculty
pstone [at] cs utexas edu
Areas of Interest
Other Areas
Reinforcement Learning
Simulated Robot Soccer
Labs
Learning Agents