Graph Neural Network Sparsification

This repository contains the code supporting the research presented in the report.

In this work, we explore methods to sparsify graph-based datasets for efficient training and inference in graph neural networks (GNNs) without compromising performance. We propose an improved algorithm for finding graph lottery tickets that yield competitive results with fewer edges.

Getting Started

Prerequisites

To set up and run the code, ensure you have the following installed:

• Python 3.13
• Libraries: PyTorch and other dependencies listed in requirements.txt

Installation

1. Clone the repository:

   git clone https://github.com/YBaumann/Python_Graphs_Test.git
   cd Python_Graphs_Test

2. Install requirements:

    pip install -r requirements.txt

3. Create a new folder:

    mkdir output_csv/