Regstar: Efficient Strategy Synthesis for Adversarial Patrolling Games
Abstract
We design a new efficient strategy synthesis method applicable to adversarial patrolling problems on graphs with arbitrary-length edges and possibly imperfect intrusion detection. The core ingredient is an efficient algorithm for computing the value and the gradient of a function assigning to every strategy its “protection” achieved. This allows for designing an efficient strategy improvement algorithm by differentiable programming and optimization techniques. Our method is the first one applicable to real-world patrolling graphs of reasonable sizes. It outperforms the state-of-the-art strategy synthesis algorithm by a margin.
Cite
Text
Klaška et al. "Regstar: Efficient Strategy Synthesis for Adversarial Patrolling Games." Uncertainty in Artificial Intelligence, 2021.Markdown
[Klaška et al. "Regstar: Efficient Strategy Synthesis for Adversarial Patrolling Games." Uncertainty in Artificial Intelligence, 2021.](https://mlanthology.org/uai/2021/klaska2021uai-regstar/)BibTeX
@inproceedings{klaska2021uai-regstar,
title = {{Regstar: Efficient Strategy Synthesis for Adversarial Patrolling Games}},
author = {Klaška, David and Kučera, Antonín and Musil, Vít and Řehák, Vojtěch},
booktitle = {Uncertainty in Artificial Intelligence},
year = {2021},
pages = {471-481},
volume = {161},
url = {https://mlanthology.org/uai/2021/klaska2021uai-regstar/}
}