zhuyiche / DeepADoTS

Repository of the paper "A Systematic Evaluation of Deep Anomaly Detection Methods for Time Series".

Anomaly Detection on Time Series: An Evaluation of Deep Learning Methods.

The goal of this repository is to provide a benchmarking pipeline for anomaly detection on time series data for multiple state-of-the-art deep learning methods.

Authors/Contributors

Team:

• Maxi Fischer
• Willi Gierke
• Thomas Kellermeier
• Ajay Kesar
• Axel Stebner
• Daniel Thevessen

Supervisors:

• Lukas Ruff
• Fabian Geier
• Emmanuel Müller

Implemented Algorithms

LSTM-AD

Malhotra, Pankaj, et al. "Long short term memory networks for anomaly detection in time series." Proceedings. Presses universitaires de Louvain, 2015.

LSTM-ED

Malhotra, Pankaj, et al. "LSTM-based encoder-decoder for multi-sensor anomaly detection." ICML, 2016.

Autoencoder

Hawkins, Simon, et al. "Outlier detection using replicator neural networks." DaWaK, 2002.

Donut

Xu, Haowen, et al. "Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applications." WWW, 2018.

REBM using a restricted Boltzmann Machine as energy-based model

Zhai, Shuangfei, et al. "Deep structured energy based models for anomaly detection." ICML, 2016.

DAGMM

Zong, Bo, et al. "Deep autoencoding gaussian mixture model for unsupervised anomaly detection." ICLR, 2018.

LSTM-DAGMM

Extension of Dagmm using an LSTM-Autoencoder instead of a Neural Network Autoencoder

Installation

git clone git://github.com/KDD-OpenSource/DeepADoTS.git  
virtualenv venv -p /usr/bin/python3  
source venv/bin/activate  
pip install -r requirements.txt

Usage

In the local repository folder, activate a virtual environment first

source venv/bin/activate
python3 main.py

Deployment

• You can use nvidia-docker
• docker build -t deep-adots .
• nvidia-docker run -ti deep-adots /bin/bash -c "python3.6 /repo/main.py"

Credits

Base implementation for DAGMM
Base implementation for Donut
Base implementation for Recurrent EBM