Bayesian Reinforcement Learning via Deep, Sparse Sampling
Divya Grover, Debabrota Basu, Christos Dimitrakakis
Abstract
We address the problem of Bayesian reinforcement learning using efficient model-based online planning. We propose an optimism-free Bayes-adaptive algorithm to induce deeper and sparser exploration with a theoretical bound on its performance relative to the Bayes optimal as well as lower computational complexity. The main novelty is the use of a candidate policy generator, to generate long-term options in the planning tree (over beliefs), which allows us to create much sparser and deeper trees. Experimental results on different environments show that in comparison to the state-of-the-art, our algorithm is both computationally more efficient, and obtains significantly higher reward over time in discrete environments.
BibTeX
@InProceedings{pmlr-v108-grover20a,
title = {Bayesian Reinforcement Learning via Deep, Sparse Sampling},
author = {Grover, Divya and Basu, Debabrota and Dimitrakakis, Christos},
booktitle = {Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics},
pages = {3036--3045},
year = {2020},
editor = {Chiappa, Silvia and Calandra, Roberto},
volume = {108},
series = {Proceedings of Machine Learning Research},
month = {26--28 Aug},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v108/grover20a/grover20a.pdf},
url = {https://proceedings.mlr.press/v108/grover20a.html},
abstract = { We address the problem of Bayesian reinforcement learning using efficient model-based online planning. We propose an optimism-free Bayes-adaptive algorithm to induce deeper and sparser exploration with a theoretical bound on its performance relative to the Bayes optimal as well as lower computational complexity. The main novelty is the use of a candidate policy generator, to generate long-term options in the planning tree (over beliefs), which allows us to create much sparser and deeper trees. Experimental results on different environments show that in comparison to the state-of-the-art, our algorithm is both computationally more efficient, and obtains significantly higher reward over time in discrete environments.}
}