Efficient Planning Under Uncertainty with Incremental Refinement

Juan Carlos Saborío, Joachim Hertzberg
Proceedings of The 35th Uncertainty in Artificial Intelligence Conference, PMLR 115:303-312, 2020.

Abstract

Online planning under uncertainty on robots and similar agents has very strict performance requirements in order to achieve reasonable behavior in complex domains with limited resources. The underlying process of decision-making and information gathering is correctly modeled by POMDP’s, but their complexity makes many interesting and challenging problems virtually intractable. We address this issue by introducing a method to estimate relevance values for elements of a planning domain, that allow an agent to focus on promising features. This approach reduces the effective dimensionality of problems, allowing an agent to plan faster and collect higher rewards. Experimental validation was performed on two challenging POMDP’s that resemble real-world robotic task planning, where it is crucial to interleave planning and acting in an efficient manner.

Cite this Paper

BibTeX
@InProceedings{pmlr-v115-saborio20a, title = {Efficient Planning Under Uncertainty with Incremental Refinement}, author = {Sabor{\'{i}}o, Juan Carlos and Hertzberg, Joachim}, booktitle = {Proceedings of The 35th Uncertainty in Artificial Intelligence Conference}, pages = {303--312}, year = {2020}, editor = {Adams, Ryan P. and Gogate, Vibhav}, volume = {115}, series = {Proceedings of Machine Learning Research}, month = {22--25 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v115/saborio20a/saborio20a.pdf}, url = {https://proceedings.mlr.press/v115/saborio20a.html}, abstract = {Online planning under uncertainty on robots and similar agents has very strict performance requirements in order to achieve reasonable behavior in complex domains with limited resources. The underlying process of decision-making and information gathering is correctly modeled by POMDP’s, but their complexity makes many interesting and challenging problems virtually intractable. We address this issue by introducing a method to estimate relevance values for elements of a planning domain, that allow an agent to focus on promising features. This approach reduces the effective dimensionality of problems, allowing an agent to plan faster and collect higher rewards. Experimental validation was performed on two challenging POMDP’s that resemble real-world robotic task planning, where it is crucial to interleave planning and acting in an efficient manner.} }
Endnote
%0 Conference Paper %T Efficient Planning Under Uncertainty with Incremental Refinement %A Juan Carlos Saborío %A Joachim Hertzberg %B Proceedings of The 35th Uncertainty in Artificial Intelligence Conference %C Proceedings of Machine Learning Research %D 2020 %E Ryan P. Adams %E Vibhav Gogate %F pmlr-v115-saborio20a %I PMLR %P 303--312 %U https://proceedings.mlr.press/v115/saborio20a.html %V 115 %X Online planning under uncertainty on robots and similar agents has very strict performance requirements in order to achieve reasonable behavior in complex domains with limited resources. The underlying process of decision-making and information gathering is correctly modeled by POMDP’s, but their complexity makes many interesting and challenging problems virtually intractable. We address this issue by introducing a method to estimate relevance values for elements of a planning domain, that allow an agent to focus on promising features. This approach reduces the effective dimensionality of problems, allowing an agent to plan faster and collect higher rewards. Experimental validation was performed on two challenging POMDP’s that resemble real-world robotic task planning, where it is crucial to interleave planning and acting in an efficient manner.
APA
Saborío, J.C. & Hertzberg, J.. (2020). Efficient Planning Under Uncertainty with Incremental Refinement. Proceedings of The 35th Uncertainty in Artificial Intelligence Conference, in Proceedings of Machine Learning Research 115:303-312 Available from https://proceedings.mlr.press/v115/saborio20a.html.

Related Material