Thanks to the support of emergenCity I received a short-term financial grant to participate and conduct research and collaborate with conference delegates at the IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP) in Costa Rica, December 2023.

I received the opportunity to present my work, “Solving FDR-Controlled Sparse Regression Problems With Five Million Variables on a Laptop” within the session for “Computational Advances in Signal Processing Algorithms”. My contribution focused on improvements of the T-Rex Selector’s computational tractability by means of memory-mapping, dummy permutations, and online-processing, and made it possible to use the T-Rex Selector for problems with millions of variables and False-Discovery Rate Control not just on a cluster computer, but also on private machines.

Furthermore, the conference granted me the opportunity to meet and discuss with domain experts and set up contacts among world leading scientists. The conference visit was part of my ongoing PhD project in track of the BMBF cluster for Future Initiative: curATime, in collaboration with the University Medical Center of Mainz.

Next to academic discussions and networking, this trip also made me a cultural expert for Costa Rica and its magnificent nature. Many thanks to emergenCity – without you it would not have been possible!

Author:

Fabian Scheidt is a scientist at TU Darmstadt in the Signal Processing Group and the Robust Data Science Group at the Institute of Telecommunications at TU Darmstadt. (updated October 24, 2024)