Graph Convolutional Networks (GCN) in PyTorch

• This is an implementation of Graph Convolutional Networks (GCNs) for node classification on the Cora citation dataset.

---

Setup

Before running the code, make sure you have installed all required dependencies. You can easily install them using the provided 'requirements.txt' file:

pip install -r requirements.txt

This code is tested with Python 3.11.9

---

Dataset: Cora

• Nodes: Scientific publications
  Each node represents an academic paper.

• Edges: Citation relationships
  An edge exists between two papers if one cites the other.

• Node Features: Word vectors
  Each publication is represented by a 1,433-dimensional bag-of-words feature vector.

• Labels (7 Classes): Publication subjects
  The Cora dataset contains 2,708 papers classified into 7 research categories.

---

Dataset Summary

Property	Description
Number of Nodes	2,708
Number of Edges	5,429
Number of Features per Node	1,433
Number of of Classes	7
Graph Type	Citation Network

---

Overview

This project demonstrates how Graph Convolutional Networks (GCNs) can learn node representations by leveraging both graph structure and node features.
The objective is to classify each publication into its corresponding research topic based on citation relationships.

---

Reference

Thomas N. Kipf and Max Welling.
Semi-Supervised Classification with Graph Convolutional Networks.
ICLR 2017. [Paper Link]

---