https://github.com/juliapomdp/factoredvaluemcts.jl
Scalable MCTS for team scenarios
https://github.com/juliapomdp/factoredvaluemcts.jl
mcts multiagent-planning multiagent-systems
Last synced: 2 months ago
JSON representation
Scalable MCTS for team scenarios
- Host: GitHub
- URL: https://github.com/juliapomdp/factoredvaluemcts.jl
- Owner: JuliaPOMDP
- License: mit
- Created: 2020-10-29T22:27:26.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2024-06-14T01:51:13.000Z (11 months ago)
- Last Synced: 2025-01-20T18:12:25.689Z (3 months ago)
- Topics: mcts, multiagent-planning, multiagent-systems
- Language: Julia
- Homepage:
- Size: 273 KB
- Stars: 15
- Watchers: 10
- Forks: 3
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
- Citation: CITATION.bib
Awesome Lists containing this project
README
# FactoredValueMCTS
[](https://github.com/JuliaPOMDP/FactoredValueMCTS.jl/actions/workflows/ci.yml)
[](http://codecov.io/github/JuliaPOMDP/FactoredValueMCTS.jl?branch=master)
[](https://juliapomdp.github.io/FactoredValueMCTS.jl/stable)
[](https://juliapomdp.github.io/FactoredValueMCTS.jl/dev)
This package implements the Monte Carlo Tree Search (MCTS) planning algorithm for Multi-Agent MDPs. The algorithm factorizes the true action value function, based on the locality of interactions between agents that is encoded with a Coordination Graph. We implement two schemes for coordinating the actions for the team of agents during the MCTS computations. The first is the iterative message-passing MaxPlus, while the second is the exact Variable Elimination. We thus get two different Factored Value MCTS algorithms, FV-MCTS-MaxPlus and FV-MCTS-VarEl respectively.
The full FV-MCTS-MaxPlus algorithm is described in our AAMAS 2021 paper _Scalable Anytime Planning for Multi-Agent MDPs_ ([Arxiv](https://arxiv.org/abs/2101.04788)). The FV-MCTS-Varel is based on the Factored Statistics algorithm from the AAAI 2015 paper _Scalable Planning and Learning from Multi-Agent POMDPs_ ([Extended Version](https://arxiv.org/abs/1404.1140)) applied to Multi-Agent MDPs rather than POMDPs. We use the latter as a baseline and show how the former outperforms it on two distinct simulated domains.
To use our solver, the domain must implement the interface from [MultiAgentPOMDPs.jl](https://github.com/JuliaPOMDP/MultiAgentPOMDPs.jl). For examples, please see [MultiAgentSysAdmin](https://github.com/JuliaPOMDP/MultiAgentSysAdmin.jl) and [MultiUAVDelivery](https://github.com/JuliaPOMDP/MultiUAVDelivery.jl), which are the two domains from our AAMAS 2021 paper. Experiments from the paper are available at https://github.com/rejuvyesh/FVMCTS_experiments.
## Installation
```julia
using Pkg
Pkg.add("FactoredValueMCTS")
```
## Citation
```
@inproceedings{choudhury2021scalable,
title={Scalable Anytime Planning for Multi-Agent {MDP}s},
author={Shushman Choudhury and Jayesh K Gupta and Peter Morales and Mykel J Kochenderfer},
booktitle={International Conference on Autonomous Agents and MultiAgent Systems},
year={2021}
}
```