Implementation code for the paper "Graph Neural Network-Based Anomaly Detection in Multivariate Time Series" (AAAI 2021)

"Rethinking Graph Neural Networks for Anomaly Detection" in ICML 2022

Code implementation for : [Graph Neural Network-Based Anomaly Detection in Multivariate Time Series(AAAI'21)](https://arxiv.org/pdf/2106.06947.pdf)

An AI-based system utilizing Graph Neural Networks (GNNs) for real-time anomaly detection and fault diagnosis in spacecraft engines. It classifies anomalies and provides actionable recommendations, improving safety and predictive maintenance.

Code Repository for Paper "HRGCN: Heterogeneous Graph-level Anomaly Detection with Hierarchical Relation-augmented Graph Neural Networks"

Anomaly detection algorithm for social networks using Graph Neural Networks by leveraging graph parameteres, between centrality, degree, closeness, on Enron and Twitter datasets

GAMA: A Multi-graph-based Anomaly Detection Framework for Business Processes via Graph Neural Networks

test video of the proposed method in "A Novel Anomaly Detection Method for Multimodal WSN Data Flow via a Dynamic Graph Neural Network"

Controlled graph neural networks with denoising diffusion for anomaly detection, Expert Systems with Applications 2023

Code and data for replicating the work in ‘Towards deep probabilistic graph neural network for natural gas leak detection and localization without labeled anomaly data’ by Zhang et al. (2023). The paper proposes a deep probabilistic graph neural network for modeling spatial sensor dependency and localizing natural gas leaks.

A project using Graph Neural Networks (GNNs) to detect and classify illicit Bitcoin transactions in the Elliptic Dataset. It explores various GNN architectures and graph-level analytics for financial anomaly detection.

Graph Neural Networks and Transformers for Jet Classification from the paper "Graph theory inspired anomaly detection at the LHC": https://arxiv.org/abs/2506.19920

Project for Columbia University COMSW4995 - Deep Learning.In this project we will use spatial-temporal graph neural networks (STGNNs) for anomaly detection on observed local area network traffic, with the purpose of detecting network intrusion attacks.

This project enhances an LSTM autoencoder for IoT anomaly detection by incorporating a Graph Neural Network (GNN) to capture the topological relationships between sensors. The model is deployed as a Containerd-based Over-the-Air (OTA) image optimized for edge devices.

Graph Neural Network Test Environment for Anomaly Detection in Provenance Graph Data

DGAD: Real-time anomaly detection in dynamic graphs using Graph Neural Networks (GNNs). Ideal for cybersecurity, fraud detection, and network monitoring.

Multi-task trained Graph Neural Network for Business Process Anomaly Detection with a Limited Number of Labeled Anomalies

A comprehensive research platform for Graph Neural Networks (GNNs) featuring 10+ applications including node and graph classification, link prediction, community detection, anomaly detection, and dynamic graph modelling, all with an interactive web interface.

AI-based fraud and anomaly detection prototype for government spending using graph neural networks and hybrid risk scoring to flag high-risk transactions for early investigation and improved transparency.

DAGNN: Deep Autoencoder-based Graph Neural Network for Local Anomaly Detection

Integrating Graph Neural Networks for Early Detection and Anomaly Identification of Acute Kidney Injury Patients

Granomaly: A Framework for Anomaly Detection in 5G Core Network Control Plane Traffic with Temporal Graph Neural Networks

Graph Neural Network-based Anomaly Detection for User Behavior in Data Systems: GCN | GAT | GAE

Official implementation of W-HGAD: a Wasserstein-based heterogeneous graph neural network for uncertainty-aware anomaly detection on graphs.

Code and data for Graph Neural Network based Robust Anomaly Detection at Service Level in SDN Driven Microservice System

This project implements an Intrusion Detection System (IDS) for smart city networks using Graph Neural Networks. By modeling IoT and cyber-physical infrastructures as graphs, the system leverages GNNs to detect anomalies, cyberattacks, and malicious traffic patterns with high accuracy.

Design and prototype two advanced anomaly detection approaches for fraud detection — a Graph Neural Network (GNN) model and a Variational Autoencoder (VAE) model — and benchmark them against a traditional gradient-boosted baseline.

LogGraph-SSL is a parsing-free anomaly detection framework for distributed system logs using Graph Neural Networks (GNNs) and Self-Supervised Learning (SSL). The framework builds token co-occurrence graphs from raw log messages without requiring manual log parsing or templates.

Credit Card Fraud Detection is Anomaly Detection Model which take in list of credit card transactions and predict whether the transaction is fraudulent or legitimate. Our model used Kaggle dataset including transactions that occurred in two days, where we have 492 frauds out of 284,807 transactions. The dataset is highly unbalanced, the positive class (frauds) account for 0.172% of all transactions Building a Computational Graph using Linear Regression Model and building functions using Neural Networks to Test/Train the model for more Accuracy and Precision.

A comprehensive, production-ready implementation of anomaly detection in graphs using various Graph Neural Network architectures. This project provides multiple models, evaluation metrics, and an interactive demo for exploring anomaly detection results.