Präsentation
Filtern
Erscheinungsjahr
- 2021 (2)
Dokumenttyp
- Vortrag (1)
- Posterpräsentation (1)
Sprache
- Deutsch (2) (entfernen)
Referierte Publikation
- nein (2)
Schlagworte
- Machine learning (2) (entfernen)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (1)
Sensors can fail. Redundancy should therefore be a design driver of wireless sensor networks. For a sensor network deployed in a steel factory, we analyze the correlations between sensors and build machine learning forecasting models, to investigate how well the network can compensate for the outage of sensors.
The stability of materials is a very important property for materials discovery. In recent years, we have explored several approaches for advancing the prediction of the stability of crystals (e.g., classical chemical heuristics and machine-learned interatomic potentials). For example, we have assessed a famous chemical heuristic – the Pauling rules – regarding their usefulness for structure prediction. They are only of limited predictive power, i.e. the rules 2 to 5 only work for 13 % of all tested oxides. Furthermore, we have shown that machine-learned interatomic potentials can be used to predict phonon properties of a range of silicon allotropes accurately and, therefore, they can also be used to assess the dynamic stability of materials. To do so, we have developed new schemes to build reference databases for machine learned interatomic potentials.