The workshop program “Machine Learning in Combustion”, offered by the research project NHR4CES, introduces students to the opportunities and potential of machine learning in combustion modeling. Through hands-on Python exercises, participants build real working examples of neural network architectures and gain an understanding of both the possibilities and the limitations of these methods.
In the past, such training sessions relied on students’ personal laptops – bringing with them a wide variety of operating systems and software configurations. Preparing every device for the course often required significant support from the supervisors. This year, however, the workshop took a step forward. With around 30 participants, the hands-on session on March 6th was successfully moved to the NFDI4ING Jupyter Service.
This online platform provides a consistent and interoperable software environment as Jupyter notebooks. The students access it through their web browsers, regardless of their personal setups, and begin coding immediately. “Using the NFDI4ING service helped us a lot. We spent far less time on technical issues and could truly focus on the training,” says Nicolas Eckel, one of the event organizers from NHR4CES SDL Energy Conversion. Special thanks go to Anett Seeland from the Jupyter Service, who prepared the software environment and ensured accessibility for all participants.
Building on this success, future training sessions will continue to use the Jupyter Service for seamless learning experiences.
Johannes Mich, STFS TU Darmstadt