|Title:||Incorporating Uncertainty into Unsupervised Machine Learning for Cyber-Physical Systems||Authors:||Voß, Carlo
|Language:||en||Keywords:||Universitätsbibliographie;Evaluation 2020||Issue Date:||10-Jun-2020||Publisher:||IEEE||Document Type:||Conference Object||Source:||Enthalten in: 2020 IEEE Conference on Industrial Cyberphysical Systems (ICPS). - Piscataway, NJ : IEEE, 2020. - 1 Online-Ressource . - 2020, Seite 475-480||Page Start:||475||Page End:||480||Pages:||475 - 480||Publisher Place:||Piscataway, NJ||Conference:||3rd IEEE International Conference on Industrial Cyber-Physical Systems||Abstract:||
© 2020 IEEE. In the field of Cyber-Physical Systems (CPS), the early detection of anomalies is crucial to avoid future faulty behaviors, e.g. preventing downtimes or decreasing product qualities. As a solution, unsupervised machine learning can be used to learn models of the historic system behavior and consequentially detect deviations from these models. Since CPS data are high dimensional time series, suitable approaches such as Long Short-Term Memory (LSTM) neural networks are good solution candidates.CPSs have specific requirements for such machine learning algorithms. Learned models must be especially useful in closed control loops, i.e. without human supervision. For this, it is essential that the uncertainty about model predictions is also part of the learned models. In order to incorporate such uncertainties, we modify the mean squared error loss function used by LSTM. This paper also analyses the solution on artificial and real data.
|Organization Units (connected with the publication):||Informatik im Maschinenbau||URL:||https://ub.hsu-hh.de/DB=1.8/XMLPRS=N/PPN?PPN=1756961948
|Appears in Collections:||2020|
Show full item record
checked on May 29, 2022
Items in openHSU are protected by copyright, with all rights reserved, unless otherwise indicated.