Machine Learning Models in Industrial Domain Changing Environments

Location Tagungshotel NH Collection Nürnberg
Anbieter Marcel Rothering
Termin

Do  16.04.   11:30-12:30 Uhr (60Min.)
Event-iCal

Teilnehmerzahl Max. 60 Teilnehmer
Rollstuhl Diese Veranstaltung ist rollstuhltauglich
-

Marcel Rothering

ML models are often built on a static extract (historical data) of actual evolving data. Especially in industrial processes, data may change rapidly and sometimes even in a hidden manner. There are several reasons for such changes of the input data. For instance, recalibration of sensors, maintenance work on machines, software and hardware upgrades of the production line or changes in the product portfolio. The provider of the ML application will not always be informed about the temporary or small changes on the production line. Certainly, not all changes will influence the model dramatically, but some of them or their combination might lead to serious consequences.

Der Vortrag wird auf Englisch gehalten.

Marcel Rothering

Education:

  • 10/2007 – 10/2010: Bachelor of Science in Physics at the Westfälische Wilhelms-Universität Münster; specialized in theoretical solid state physics
  • 10/2010 – 12/2012: Master of Science in Physics at the Westfälische Wilhelms-Universität Münster; specialized in theoretical particle physics
  • 12/2012 – 12/2016: PhD in Physics at the Westfälische Wilhelms-Universität Münster; Topic: Precise Predictions for Supersymmetric Particle Production at the LHC

Work experience:

  • 10/2012 – 12/2016: Research Scientist at the Institute for theoretical physics at the Westfälische Wilhelms-Universität Münster
  • 02/2017 – 12/2018: IT Consultant at saracus consulting GmbH in Münster; focus on data driven projects for insurance companies
  • 03/2019 – present: Data Scientist at Siemens AG


Das Foto wurde uns vom Referenten freundlicherweise für diese Webseite zur Verfügung gestellt.

EventID: 224