Anomaly Detection for Water Treatment System based on Neural Network with Automatic Architecture Optimization

2018/07/19 by Dmitry Shalyga, Shalyga, Dmitry, Pavel Filonov +3 · 1 citation
Computer Science · Engineering · #Anomaly Detection Techniques and Applications #FOS: Computer and information sciences #Fault Detection and Control Systems #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Network Security and Intrusion Detection

paper · pdf · doi:10.48550/arxiv.1807.07282

openalex publication_date 2018/07/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We continue to develop our neural network (NN) based forecasting approach to anomaly detection (AD) using the Secure Water Treatment (SWaT) industrial control system (ICS) testbed dataset. We propose genetic algorithms (GA) to find the best NN architecture for a given dataset, using the NAB metric to assess the quality of different architectures. The drawbacks of the F1-metric are analyzed. Several techniques are proposed to improve the quality of AD: exponentially weighted smoothing, mean p-powered error measure, individual error weight for each variable, disjoint prediction windows. Based on the techniques used, an approach to anomaly interpretation is introduced.

Citations

• Gradient-based learning applied to document recognition

Cited by

• Anomaly Detection for Industrial Control Systems Using Sequence-to-Sequence Neural Networks

Related

• Anomaly Detection for Industrial Control Systems Using Sequence-to-Sequence Neural Networks
• Detection of Anomalies and Faults in Industrial IoT Systems by Data Mining: Study of CHRIST Osmotron Water Purification System
• Detecting Cyberattacks in Industrial Control Systems Using Convolutional Neural Networks
• MAD-GAN: Multivariate Anomaly Detection for Time Series Data with Generative Adversarial Networks
• Graph Neural Network-Based Anomaly Detection for River Network Systems