Action Schema Networks: Generalised Policies with Deep Learning

Abstract

In this paper, we introduce the Action Schema Network (ASNet): a neural network architecture for learning generalised policies for probabilistic planning problems. By mimicking the relational structure of planning problems, ASNets are able to adopt a weight sharing scheme which allows the network to be applied to any problem from a given planning domain. This allows the cost of training the network to be amortised over all problems in that domain. Further, we propose a training method which balances exploration and supervised training on small problems to produce a policy which remains robust when evaluated on larger problems. In experiments, we show that ASNet's learning capability allows it to significantly outperform traditional non-learning planners in several challenging domains.

Cite

Text

Toyer et al. "Action Schema Networks: Generalised Policies with Deep Learning." AAAI Conference on Artificial Intelligence, 2018. doi:10.1609/AAAI.V32I1.12089

Markdown

[Toyer et al. "Action Schema Networks: Generalised Policies with Deep Learning." AAAI Conference on Artificial Intelligence, 2018.](https://mlanthology.org/aaai/2018/toyer2018aaai-action/) doi:10.1609/AAAI.V32I1.12089

BibTeX

@inproceedings{toyer2018aaai-action,
  title     = {{Action Schema Networks: Generalised Policies with Deep Learning}},
  author    = {Toyer, Sam and Trevizan, Felipe W. and Thiébaux, Sylvie and Xie, Lexing},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2018},
  pages     = {6294-6301},
  doi       = {10.1609/AAAI.V32I1.12089},
  url       = {https://mlanthology.org/aaai/2018/toyer2018aaai-action/}
}