Generative Adversarial Imitation Learning from Failed Experiences (Student Abstract)

Jiacheng Zhu, Jiahao Lin, Meng Wang, Yingfeng Chen, Changjie Fan, Chong Jiang, Zongzhang Zhang

AAAI 2020 pp. 13997-13998

doi:10.1609/AAAI.V34I10.7271 /aaai/2020/zhu2020aaai-generative/

Abstract

Imitation learning provides a family of promising methods that learn policies from expert demonstrations directly. As a model-free and on-line imitation learning method, generative adversarial imitation learning (GAIL) generalizes well to unseen situations and can handle complex problems. In this paper, we propose a novel variant of GAIL called GAIL from failed experiences (GAILFE). GAILFE allows an agent to utilize failed experiences in the training process. Moreover, a constrained optimization objective is formalized in GAILFE to balance learning from given demonstrations and from self-generated failed experiences. Empirically, compared with GAIL, GAILFE can improve sample efficiency and learning speed over different tasks.

PDF AAAI Semantic Scholar

Cite

Text

Zhu et al. "Generative Adversarial Imitation Learning from Failed Experiences (Student Abstract)." AAAI Conference on Artificial Intelligence, 2020. doi:10.1609/AAAI.V34I10.7271

Markdown

[Zhu et al. "Generative Adversarial Imitation Learning from Failed Experiences (Student Abstract)." AAAI Conference on Artificial Intelligence, 2020.](https://mlanthology.org/aaai/2020/zhu2020aaai-generative/) doi:10.1609/AAAI.V34I10.7271

BibTeX

@inproceedings{zhu2020aaai-generative,
  title     = {{Generative Adversarial Imitation Learning from Failed Experiences (Student Abstract)}},
  author    = {Zhu, Jiacheng and Lin, Jiahao and Wang, Meng and Chen, Yingfeng and Fan, Changjie and Jiang, Chong and Zhang, Zongzhang},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2020},
  pages     = {13997-13998},
  doi       = {10.1609/AAAI.V34I10.7271},
  url       = {https://mlanthology.org/aaai/2020/zhu2020aaai-generative/}
}