| Time: | February 9, 2022, 2:00 p.m. – 4:00 p.m. |
|---|---|
| Download as iCal: |
|
The Special Interest Group Data Infrastructure offers a forum to interested working groups that want to set up or further develop an RDM infrastructure at working group or institute level. We invite you to a monthly SIGDIUS seminar, to which we invite internal and external experts for presentations and discussions. SIGDIUS members will have the opportunity to exchange their experiences with concrete RDM infrastructures.
We cordially invite all interested parties to our next meeting on 9 February 2022 at 2pm. Due to the current situation, this seminar will be held as an online seminar. For participation, please send an e-mail to Juergen.Pleiss@itb.uni-stuttgart.de with the subject line "SIGDIUS seminar 9.2."
Caterina Barillari (ETH Zürich):
"Research Data Management with openBIS"
Jakob Dürrwächter (IAG/SimTech):
"An open-source fluid dynamics simulation database for validation and machine learning"
High-quality data is the most important pre-requisite for machine learning. In computational fluid dynamics, large amounts of data are generated, but the data size makes sharing difficult and expensive. In the present project, an open access fluid dynamics data base is created. It provides input to machine learning applications and benchmark data for validation of other codes. Test cases of different size and complexity are included. Sharing full time-resolved field data is unfeasible, and the data might not match the exact requirements of users. So instead, an approach of effortless full reproducibility is followed: The complete source code, build configuration and simulation setup is automatically tracked and included with every computation, and scripts are provided to recreate the simulations at the push of a button. Containers ensure system flexibility. This strategy will also be employed for simulations in future publications. It introduces a comprehensive open data and open-source paradigm, which is necessary for collaboration and transparency, but is still seldomly found in the literature.
More about the SIGDIUS seminars