A Physics-Based Model Prior for Object-Oriented MDPs
Abstract
One of the key challenges in using reinforcement learning in robotics is the need for models that capture natural world structure. There are, methods that formalize multi-object dynamics using relational representations, but these methods are not sufficiently compact for real-world robotics. We present a physics-based approach that exploits modern simulation tools to efficiently parameterize physical dynamics. Our results show that this representation can result in much faster learning, by virtue of its strong but appropriate inductive bias in physical environments.
Cite
Text
Scholz et al. "A Physics-Based Model Prior for Object-Oriented MDPs." International Conference on Machine Learning, 2014.Markdown
[Scholz et al. "A Physics-Based Model Prior for Object-Oriented MDPs." International Conference on Machine Learning, 2014.](https://mlanthology.org/icml/2014/scholz2014icml-physicsbased/)BibTeX
@inproceedings{scholz2014icml-physicsbased,
title = {{A Physics-Based Model Prior for Object-Oriented MDPs}},
author = {Scholz, Jonathan and Levihn, Martin and Isbell, Charles and Wingate, David},
booktitle = {International Conference on Machine Learning},
year = {2014},
pages = {1089-1097},
volume = {32},
url = {https://mlanthology.org/icml/2014/scholz2014icml-physicsbased/}
}