Go-GCN is an implementation of Graph Convolutional Networks (GCN) in Go. It allows for semi-supervised learning on graph-structured data, such as cita

Search code, repositories, users, issues, pull requests...

submited by

Style Pass

2024-09-23 15:00:04

Go-GCN is an implementation of Graph Convolutional Networks (GCN) in Go. It allows for semi-supervised learning on graph-structured data, such as citation networks. The project is designed with modular components for easy customization and experimentation, including support for dropout, Adam optimizer, and accuracy evaluation.

This implementation is based on the paper Semi-Supervised Classification with Graph Convolutional Networks by Thomas Kipf and Max Welling.

The project is configured to run on the Cora dataset, located in the datasets/cora directory. Use the following command to run the model:

The project uses the Cora dataset, a common benchmark for graph-based learning tasks. The dataset is located in the datasets/cora directory:

The LoadData function in internal/data/load.go is responsible for loading and processing the dataset into adjacency matrices and feature matrices.

The GCN model is implemented in internal/model/gcn.go. It supports multiple layers, dropout regularization, and training with either the Adam or SGD optimizer.