SkillAgentSearch skills...

GraphEmbeddings

Python implementation of the DDoS (“Histogram loss”) graph node embedding algorithm, with experiment pipelines for link prediction, node classification, and clustering. Paper: https://arxiv.org/abs/1810.03032

Install / Use

/learn @premolab/GraphEmbeddings
About this skill

Quality Score

0/100

Supported Platforms

Universal

README

GraphEmbeddings

This repository contains realization of DDoS (aka Histogram loss) algorithm for generating graph embeddings.

Structure

All code is stored in folder final_src

Realization of several embedding algorithm including our algorithm can be found in folder transformers.

Folder io_utils contains code responsible for reading and writing graphs and embeddings.

Folder transformation contains generic code to generate an embedding with any available algorithm.

Other folders represent sets of experiments for comparing algorithms: link_prediction, classification and clusterization.

How to run

To make this code work you need to replace path to necessary dataset in file final_src/settings.py with your local path.

View on GitHub
GitHub Stars9
CategoryDevelopment
Updated17d ago
Forks5

Languages

Python

Security Score

70/100

Audited on Mar 11, 2026

No findings