Implementation and evaluation of Almanac (Automaton/Logic Multi-Agent Natural Actor-Critic), an algorithm for Multi-Agent Reinforcement Learning with Temporal Logic Specifications. A preliminary version of this work appeared at AAMAS-21, the code for which has been archived on the branch almanac/aamas. An updated version is forthcoming and will be pushed to the branch almanac/main. If you use or reference this work in your own publications, please use the following citation:
@inproceedings{10.5555/3463952.3464024,
author = {Hammond, Lewis and Abate, Alessandro and Gutierrez, Julian and Wooldridge, Michael},
title = {Multi-Agent Reinforcement Learning with Temporal Logic Specifications},
year = {2021},
publisher = {International Foundation for Autonomous Agents and Multiagent Systems},
address = {Richland, SC},
booktitle = {Proceedings of the 20th International Conference on Autonomous Agents and MultiAgent Systems},
pages = {583–592},
location = {Virtual Event, United Kingdom},
series = {AAMAS '21}}
If you have any questions about the paper or the code, please feel free to email me.