Multi-Resolution Exploration in Continuous Spaces
Abstract
The essence of exploration is acting to try to decrease uncertainty. We propose a new methodology for representing uncertainty in continuous-state control problems. Our approach, multi-resolution exploration (MRE), uses a hierarchical mapping to identify regions of the state space that would benefit from additional samples. We demonstrate MRE's broad utility by using it to speed up learning in a prototypical model-based and value-based reinforcement-learning method. Empirical results show that MRE improves upon state-of-the-art exploration approaches.
Cite
Text
Nouri and Littman. "Multi-Resolution Exploration in Continuous Spaces." Neural Information Processing Systems, 2008.Markdown
[Nouri and Littman. "Multi-Resolution Exploration in Continuous Spaces." Neural Information Processing Systems, 2008.](https://mlanthology.org/neurips/2008/nouri2008neurips-multiresolution/)BibTeX
@inproceedings{nouri2008neurips-multiresolution,
title = {{Multi-Resolution Exploration in Continuous Spaces}},
author = {Nouri, Ali and Littman, Michael L.},
booktitle = {Neural Information Processing Systems},
year = {2008},
pages = {1209-1216},
url = {https://mlanthology.org/neurips/2008/nouri2008neurips-multiresolution/}
}