Augmenting Policy Learning with Routines Discovered from a Single Demonstration
Name
2012.12469.pdf
Description
Accepted version
Size
1.82 MB
Format
Adobe PDF
Checksum (MD5)
5df3b6275c28dc85c5fa20c271050eb9
Author(s)
Zhao, Zelin
Gan, Chuang
Wu, Jiajun
Guo, Xiaoxiao
Tenenbaum, Joshua B
Date Issued
2021
Journal
Proceedings of the AAAI Conference on Artificial Intelligence
Publisher
Association for the Advancement of Artificial Intelligence (AAAI)
Citation
Zhao, Zelin, Gan, Chuang, Wu, Jiajun, Guo, Xiaoxiao and Tenenbaum, Joshua B. 2021. "Augmenting Policy Learning with Routines Discovered from a Single Demonstration." Proceedings of the AAAI Conference on Artificial Intelligence, 35 (12).
Version
Author's final manuscript
Abstract
Humans can abstract prior knowledge from very little data and use it to boost skill learning. In this paper, we propose routine-augmented policy learning (RAPL), which discovers routines composed of primitive actions from a single demonstration and uses discovered routines to augment policy learning. To discover routines from the demonstration, we first abstract routine candidates by identifying grammar over the demonstrated action trajectory. Then, the best routines measured by length and frequency are selected to form a routine library. We propose to learn policy simultaneously at primitive-level and routine-level with discovered routines, leveraging the temporal structure of routines. Our approach enables imitating expert behavior at multiple temporal scales for imitation learning and promotes reinforcement learning exploration. Extensive experiments on Atari games demonstrate that RAPL improves the state-of-the-art imitation learning method SQIL and reinforcement learning method A2C. Further, we show that discovered routines can generalize to unseen levels and difficulties on the CoinRun benchmark.
MIT Department
Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1609/AAAI.V35I12.17316
