
The project C1 aims for the (1) the creation of a semantic model (i.e. a knowledge graph) of copolymers based on formalized domain knowledge and to steadily improve this with additional sources including available experimental data and information provided by other subprojects; (2) training of a ML model that predicts a variety of properties using limited available experimental and formalized domain knowledge, and to steadily improve it from feedback loops (active learning) to reach a high accuracy model generalizable to other materials; (3) for the leveraging of information in the semantic model to improve performance of the ML model and (4) for the usage of the models to provide insights on how the experimental conditions (processing and synthetic routes) influence the final properties of the copolymer blends.
Prof. Dr.-Ing. Holger Ruckdäschel
Image: PrivatePrincipal Investigators
Prof. Dr.-Ing. Holger Ruckdäschel
Faculty of Engineering Science
University of Bayreuth
Universitätsstraße 30
95447 Bayreuth
Phone: +49 921 55-7470
ruckdaeschel@uni-bayreuth.de
Prof. Dr. Birgitta König-Ries
Image: Jens Meyer (University of Jena)Prof. Dr. Birgitta König-Ries
Institute of Computer Science
Friedrich Schiller University Jena
Leutragraben 1
07743 Jena
Phone: +49 3641 9-46430
birgitta.koenig-ries@uni-jena.de