README.md

HDDRL

Hierarchical Decentralized Deep Reinforcement Learning

Status: Archive (code is provided as-is, no updates expected)

Figure: Visualization of the compared architectures:

(a): Centralized architecture with one controller (C₀).

(b): Fully decentralized architecture with four different controllers (C₀, C₁, C₂, C₁).

(c): Hierarchical architecture with 1 HLC and 1 LLC.

(d): Hierarchical decentralized architecture with 1 HLC and 4 LLCs.

(e): Hierarchical decentralized architecture with 4 HLCs and 4 LLCs.

Install

Install the conda package manager from https://docs.conda.io/en/latest/miniconda.html

# Required: Sampling
conda create --name hddrl python=3.6.13
conda activate hddrl
git clone https://github.com/wzaiealmri/hddrl.git
cd hddrl
pip install -r requirements.txt
pip install -e .

Citation

@article{wzaielamri_HDDRL,
  title={Hierarchical Decentralized Deep Reinforcement Learning Architecture for a Simulated Four-Legged Agent.},
  author={Wadhah Zai El Amri and Luca Hermes and Malte Schilling},
  journal={Proc. of 8th International Online & Onsite Conference on Machine Learning, Optimization, and Data Science.},
  year={2022}
}