dc.contributor.author |
Van Niekerk, B
|
|
dc.contributor.author |
Damianou, A
|
|
dc.contributor.author |
Rosman, Benjamin S
|
|
dc.date.accessioned |
2017-09-29T06:47:28Z |
|
dc.date.available |
2017-09-29T06:47:28Z |
|
dc.date.issued |
2017-08 |
|
dc.identifier.citation |
Van Niekerk, B., Damianou, A. and Rosman, B.S. 2017. Online constrained model-based reinforcement learning. Uncertainty in Artificial Intelligence, 11-15 August 2017, Sydney, Australia |
en_US |
dc.identifier.uri |
http://auai.org/uai2017/proceedings/papers/276.pdf
|
|
dc.identifier.uri |
http://hdl.handle.net/10204/9617
|
|
dc.description |
Uncertainty in Artificial Intelligence, 11-15 August 2017, Sydney, Australia |
en_US |
dc.description.abstract |
Applying reinforcement learning to robotic systems poses a number of challenging problems. A key requirement is the ability to handle continuous state and action spaces while remaining within a limited time and resource budget. Additionally, for safe operation, the system must make robust decisions under hard constraints. To address these challenges, we propose a model based approach that combines Gaussian Process regression and Receding Horizon Control. Using sparse spectrum Gaussian Processes, we extend previous work by updating the dynamics model incrementally from a stream of sensory data. This results in an agent that can learn and plan in real-time under non-linear constraints. We test our approach on a cart pole swing-up environment and demonstrate the benefits of online learning on an autonomous racing task. The environment’s dynamics are learned from limited training data and can be reused in new task instances without retraining. |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
Association for Uncertainty in Artificial Intelligence (AUAI) |
en_US |
dc.relation.ispartofseries |
Worklist;19464 |
|
dc.subject |
Model-based reinforcement learning |
en_US |
dc.subject |
Gaussian Process regression |
en_US |
dc.subject |
Receding horizon control |
en_US |
dc.subject |
Artificial intelligence |
en_US |
dc.title |
Online constrained model-based reinforcement learning |
en_US |
dc.type |
Conference Presentation |
en_US |
dc.identifier.apacitation |
Van Niekerk, B., Damianou, A., & Rosman, B. S. (2017). Online constrained model-based reinforcement learning. Association for Uncertainty in Artificial Intelligence (AUAI). http://hdl.handle.net/10204/9617 |
en_ZA |
dc.identifier.chicagocitation |
Van Niekerk, B, A Damianou, and Benjamin S Rosman. "Online constrained model-based reinforcement learning." (2017): http://hdl.handle.net/10204/9617 |
en_ZA |
dc.identifier.vancouvercitation |
Van Niekerk B, Damianou A, Rosman BS, Online constrained model-based reinforcement learning; Association for Uncertainty in Artificial Intelligence (AUAI); 2017. http://hdl.handle.net/10204/9617 . |
en_ZA |
dc.identifier.ris |
TY - Conference Presentation
AU - Van Niekerk, B
AU - Damianou, A
AU - Rosman, Benjamin S
AB - Applying reinforcement learning to robotic systems poses a number of challenging problems. A key requirement is the ability to handle continuous state and action spaces while remaining within a limited time and resource budget. Additionally, for safe operation, the system must make robust decisions under hard constraints. To address these challenges, we propose a model based approach that combines Gaussian Process regression and Receding Horizon Control. Using sparse spectrum Gaussian Processes, we extend previous work by updating the dynamics model incrementally from a stream of sensory data. This results in an agent that can learn and plan in real-time under non-linear constraints. We test our approach on a cart pole swing-up environment and demonstrate the benefits of online learning on an autonomous racing task. The environment’s dynamics are learned from limited training data and can be reused in new task instances without retraining.
DA - 2017-08
DB - ResearchSpace
DP - CSIR
KW - Model-based reinforcement learning
KW - Gaussian Process regression
KW - Receding horizon control
KW - Artificial intelligence
LK - https://researchspace.csir.co.za
PY - 2017
T1 - Online constrained model-based reinforcement learning
TI - Online constrained model-based reinforcement learning
UR - http://hdl.handle.net/10204/9617
ER -
|
en_ZA |