Skip to content

AutoEncoders for Event Detection (AEED): a Keras-based class for anomaly detection in water sensor networks.

Notifications You must be signed in to change notification settings

Critical-Infrastructure-Systems-Lab/aeed

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

AutoEncoders for Event Detection (AEED): a Keras-based class for anomaly detection in water sensor networks.

Summary

This repository contains the code used for the following publication:
Taormina, R. and Galelli, S., 2018. Deep-Learning Approach to the Detection and Localization of Cyber-Physical Attacks on Water Distribution Systems. Journal of Water Resources Planning and Management, 144(10), p.04018065.

Although the usage is demonstrated for sensor anomalies engendered by cyber attacks on water distribution systems, the methodology could be applied for any network of sensors monitoring a physical process. AEED work by spotting incongruences in the physical relationships and spatial correlation present the recorded readings.

If you use the package, please make sure to cite the aformentioned paper, which reports all the details on the methodology.

Contents

  • aeed.py                       Python module containing the Autoencoder's Class.
  • aeed_batadal.ipynb    Python notebook demonstrating AEED usage for the problem described in the publication.
  • ./data/                        Data repository on cyber-attack scenarios on WDS, also available at www.batadal.net .

About

AutoEncoders for Event Detection (AEED): a Keras-based class for anomaly detection in water sensor networks.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 95.1%
  • Python 4.9%