Exploration by Distributional Reinforcement Learning
Abstract
We propose a framework based on distributional reinforcement learning and recent attempts to combine Bayesian parameter updates with deep reinforcement learning. We show that our proposed framework conceptually unifies multiple previous methods in exploration. We also derive a practical algorithm that achieves efficient exploration on challenging control tasks.
Cite
Text
Tang and Agrawal. "Exploration by Distributional Reinforcement Learning." International Joint Conference on Artificial Intelligence, 2018. doi:10.24963/IJCAI.2018/376Markdown
[Tang and Agrawal. "Exploration by Distributional Reinforcement Learning." International Joint Conference on Artificial Intelligence, 2018.](https://mlanthology.org/ijcai/2018/tang2018ijcai-exploration/) doi:10.24963/IJCAI.2018/376BibTeX
@inproceedings{tang2018ijcai-exploration,
title = {{Exploration by Distributional Reinforcement Learning}},
author = {Tang, Yunhao and Agrawal, Shipra},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2018},
pages = {2710-2716},
doi = {10.24963/IJCAI.2018/376},
url = {https://mlanthology.org/ijcai/2018/tang2018ijcai-exploration/}
}