Learning a state-based generative model of graph distributions.
@inproceedings{stier2020deep,
title={DeepGG: a Deep Graph Generator},
author={Stier, Julian and Granitzer, Michael},
booktitle={Advances in Intelligent Data Analysis XIX: 19th International Symposium on Intelligent Data Analysis, IDA 2021, Porto, Portugal, April 26--28, 2021, Proceedings},
pages={325},
organization={Springer Nature}
}
- install the conda environment with
conda env create -f environment.yml
- activate the environment
conda activate sur-deepgg
- configure your (hyper)parameters (first ~20 variables)
- invoke as much as possible computations via
python deepgg_pipeline.py
- merge the computations as shown in deepgg-merge-computations.ipynb
- have a look over the exemplary notebooks of how to visualize some aspects of the computed models