New Publication at SafeComp 2020 on Automated Anomaly Detection using Clustering

Paper Title: "Automated Anomaly Detection in CPS Log Files - A Time Series Clustering Approach"; Tabea Schmidt et al.

In this work, we propose to cluster the time series data of log files of Cyber-Physical Systems (CPS) to detect anomalies in the system's behavior in an automated way. Our approach is generically applicable to different kinds of CPS without having the need to initially build or learn a fine-tuned model of the system. The experimental results show that our proposed solution can effectively detect anomalies in the log files from different types of CPS.

Read the publication here: