A PyTorch implementation of "ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning", CIKM-21

Last update: Dec 14, 2022

Overview

ANEMONE

A PyTorch implementation of "ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning", CIKM-21

Dependencies

python==3.6.1
dgl==0.4.1
matplotlib==3.3.4
networkx==2.5
numpy==1.19.2
pyparsing==2.4.7
scikit-learn==0.24.1
scipy==1.5.2
sklearn==0.24.1
torch==1.8.1

To install all dependencies:

pip install -r requirements.txt

Usage

Anomalies have been injected into three datasets under the ./dataset directory. Please refer to GRAND-Lab/CoLA for graph anomaly injection.

To train and evaluate on Cora:

python run.py --expid 1 --device cuda:0 --runs 1 --alpha 0.8

To train and evaluate on Citeseer:

python run.py --dataset citeseer --expid 2 --device cuda:0 --runs 1 --alpha 0.6

To train and evaluate on Pubmed:

python run.py --dataset pubmed --expid 3 --device cuda:0 --runs 1 --alpha 0.8 --negsamp_ratio_patch 6 --negsamp_ratio_context 1

Citation

If you use our code in your research, please cite the following article:

@inproceedings{jin2021anomaly,
  title={ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning},
  author={Ming Jin and Yixin Liu and Yu Zheng and Lianhua Chi and Yuan-Fang Li and Shirui Pan},
  booktitle={Proceedings of the 30th ACM International Conference on Information \& Knowledge Management},
  year={2021}
}

A PyTorch implementation of "ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning", CIKM-21

Related tags

Overview

ANEMONE

Dependencies

Usage

Citation

Owner

Graph Analysis & Deep Learning Laboratory, GRAND

Keras Image Embeddings using Contrastive Loss

LeafSnap replicated using deep neural networks to test accuracy compared to traditional computer vision methods.

GPU implementation of $k$-Nearest Neighbors and Shared-Nearest Neighbors

A Pytorch Implementation of a continuously rate adjustable learned image compression framework.

Chainer Implementation of Semantic Segmentation using Adversarial Networks

A pytorch implementation of Detectron. Both training from scratch and inferring directly from pretrained Detectron weights are available.

A PyTorch implementation of EfficientNet and EfficientNetV2 (coming soon!)

DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort

3D-Transformer: Molecular Representation with Transformer in 3D Space

Benchmarks for the Optimal Power Flow Problem

GNPy: Optical Route Planning and DWDM Network Optimization

code for "Feature Importance-aware Transferable Adversarial Attacks"

A list of Machine Learning Art Colabs

Pytorch implementation of our paper accepted by NeurIPS 2021 -- Revisiting Discriminator in GAN Compression: A Generator-discriminator Cooperative Compression Scheme

This library provides an abstraction to perform Model Versioning using Weight & Biases.

Scikit-learn compatible estimation of general graphical models

Learn about Spice.ai with in-depth samples

Exporter for Storage Area Network (SAN)

BisQue is a web-based platform designed to provide researchers with organizational and quantitative analysis tools for 5D image data. Users can extend BisQue by implementing containerized ML workflows.

Scheduling BilinearRewards