Hanjun-Dai / steady_state_embedding

Learning Steady-States of Iterative Algorithms over Graphs

graph steady state embedding

Learning Steady-States of Iterative Algorithms over Graphs (http://proceedings.mlr.press/v80/dai18a/dai18a.pdf)

1. Setup the environment

1) Download the repository

git clone [e-mail]:Hanjun-Dai/steady_state_embedding --recursive

2) Build the dependency

This project depends on the graphnn library. The building instruction can be found here:

https://github.com/Hanjun-Dai/graphnn

3) Download the data

Use the following dropbox link:

https://www.dropbox.com/sh/3gwr2wgh455q9pi/AAB0i6EQimVGslrqtsTIWsL0a?dl=0

Put everything under the 'data' folder, or create a symbolic link with name 'data':

ln -s /path/to/your/downloaded/files data

Finally the folder structure should look like this:

steady_state_embedding (project root)
|__  README.md
|__  code
|__  graphnn
|__  data
|__  |__ algo_data
|    |__ amazon
|    |__ pagerank_ba
|......

2. Run the experiments

Most of the experiments are self-contained, where you need to build the code and then execute the script. For example:

cd code/fit_algo/connectivity
make
./connect.sh

The data is already cooked, so the code/data_process folder is for reference purpose.

Reference

@inproceedings{dai2018learning,
  title={Learning Steady-States of Iterative Algorithms over Graphs},
  author={Dai, Hanjun and Kozareva, Zornitsa and Dai, Bo and Smola, Alex and Song, Le},
  booktitle={International Conference on Machine Learning},
  pages={1114--1122},
  year={2018}
}