Machine Learning and Deep Learning Approaches for Multivariate Time Series Prediction and Anomaly Detection
- M. Thill
- Thursday 17 March 2022
2311 GJ Leiden
- Prof. T.H.W. Bäck
- Prof. W. Konen (Cologne University)
In many real-world applications today, it is critical to continuously record and monitor certain machine or system health indicators to discover malfunctions or other abnormal behavior at an early stage and prevent potential harm. The demand for such reliable monitoring systems is expected to increase in the coming years.
Particularly in the industrial context, in the course of ongoing digitization, it is becoming increasingly important to analyze growing volumes of data in an automated manner using state-of-the-art algorithms. In many practical applications, one has to deal with temporal data in the form of data streams or time series. The problem of detecting unusual (or anomalous) behavior in time series is commonly referred to as time series anomaly detection. Anomalies are events observed in the data that do not conform to the normal or expected behavior when viewed in their temporal context.
This thesis focuses on unsupervised machine learning algorithms for anomaly detection in time series. In an unsupervised learning setup, a model attempts to learn the normal behavior in a time series — which might already be contaminated with anomalies — without any external assistance. The model can then use its learned notion of normality to detect anomalous events.
Four unsupervised anomaly detection algorithms for multivariate time series are presented, which can be used in different contexts. We evaluate the algorithms presented in this work on challenging synthetic and real-world time series anomaly detection benchmarks and compare them to other state-of-the-art algorithms.
Proefschriften van Leidse promovendi zijn na de promotie digitaal beschikbaar via het Leids Repositorium. De proefschriften op deze site zijn vrij toegankelijk. Alleen in sommige gevallen rust er een tijdelijk embargo op een proefschrift en wordt het proefschrift pas later volledig beschikbaar gesteld.
Persvragen (alleen journalisten)
071 527 1521
071 527 7211