Policy Reuse in Deep Reinforcement Learning
Abstract
Driven by recent developments in Artificial Intelligence research, a promising new technology for building intelligent agents has evolved. The approach is termed Deep Reinforcement Learning and combines the classic field of Reinforcement Learning (RL) with the representational power of modern Deep Learning approaches. It is very well suited for single task learning but needs a long time to learn any new task. To speed up this process, we propose to extend the concept to multi-task learning by adapting Policy Reuse, a Transfer Learning approach from classic RL, to use with Deep Q-Networks.
Cite
Text
Glatt and Costa. "Policy Reuse in Deep Reinforcement Learning." AAAI Conference on Artificial Intelligence, 2017. doi:10.1609/AAAI.V31I1.11091Markdown
[Glatt and Costa. "Policy Reuse in Deep Reinforcement Learning." AAAI Conference on Artificial Intelligence, 2017.](https://mlanthology.org/aaai/2017/glatt2017aaai-policy/) doi:10.1609/AAAI.V31I1.11091BibTeX
@inproceedings{glatt2017aaai-policy,
title = {{Policy Reuse in Deep Reinforcement Learning}},
author = {Glatt, Ruben and Costa, Anna Helena Reali},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2017},
pages = {4929-4930},
doi = {10.1609/AAAI.V31I1.11091},
url = {https://mlanthology.org/aaai/2017/glatt2017aaai-policy/}
}