Anomaly Detection in Networks via Score-Based Generative Models

Dmitrii Gavrilev, Evgeny Burnaev
6/27/2023
cs.LG

Official code: realfolkcode/GraphDiffusionAnomaly (4 stars, last commit 2023-07-24).

Abstract

Node outlier detection in attributed graphs is a challenging problem for which there is no method that would work well across different datasets. Motivated by the state-of-the-art results of score-based models in graph generative modeling, we propose to incorporate them into the aforementioned problem. Our method achieves competitive results on small-scale graphs. We provide an empirical analysis of the Dirichlet energy, and show that generative models might struggle to accurately reconstruct it.

View on arXivView PDF

Code Implementations(1)

realfolkcode/GraphDiffusionAnomalyOfficial
40Python, Jupyter NotebookNov 13, 20223 years ago

Similar Papers

Three Revisits to Node-Level Graph Anomaly Detection: Outliers, Message Passing and Hyperbolic Neural Networks

85% match1 repo42024

GADBench: Revisiting and Benchmarking Supervised Graph Anomaly Detection

83% match1 repo1492023

Robust Anomaly Detection with Graph Neural Networks using Controllability

82% match2025

GTHNA: Local-global Graph Transformer with Memory Reconstruction for Holistic Node Anomaly Evaluation

82% match2025

GRAM: An Interpretable Approach for Graph Anomaly Detection using Gradient Attention Maps

82% match2023

Guarding Graph Neural Networks for Unsupervised Graph Anomaly Detection

82% match2024

Cite this paper

@article{gavrilev2023anomaly,
  title  = {Anomaly Detection in Networks via Score-Based Generative Models},
  author = {Dmitrii Gavrilev and Evgeny Burnaev},
  year   = {2023},
  eprint = {2306.15324},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2306.15324v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2306.15324.svg)](https://codeofpaper.com/paper/2306.15324)

Discussion