Learning on non - Euclidean data

In the framework of the OEAD program "Scientific and Technical Cooperation" (STC)  the project  "Collective evolution strategies for distributed learning on non-Euclidean data" to establish joint research activities between the research centre Business Informatics of the Vorarlberg University of Applied Sciences and the Department of Mathematics of the University of Montenegro.

Over the two-year duration, the project teams will work out common research questions in several bilateral workshops. The focus will be on the topics

• Evolutionary Algorithms

• Self-organizing complex physical systems

To combine these topics, in particular problems involving  non-Euclidean geometry shall be considered. Applications of such problems can be found for example

• in simultaneous communication between actors

• in shape optimization for the purpose of image recognition/computer graphics
• in the design of neural networks for dynamic problems

In addition to developing new methods, the collaboration will be used to prepare other joint research projects.