Jasin Machkour, scientist at emergenCITY and the Robust Data Science Group at the Institute of Communications Engineering at TU Darmstadt, received the highest distinction for his dissertation. On August 23, 2024, he defended his doctoral thesis titled “Development of Fast Machine Learning Algorithms for False Discovery Rate Control in Large-Scale High-Dimensional Data” with Summa Cum Laude. The comission was impressed by the fact that he not only worked on fundamental theoretical research questions, but also applied his methods to several challenging applications.

The core of his five years of research was the development of the “Terminating-Random Experiments Selector (T-Rex)”, an efficient method that demonstrably controls the false discovery rate in high-dimensional and sparse datasets. The method has been applied to real-world biomedical and financial problems and its open-source implementation has been downloaded more than 12,000 times to date.

Within emergenCITY, the T-Rex methods are employed to enhance the Scout robot’s ability to locate missing persons in crisis situations. The developed methods help maximize successful rescues while limiting false-positive detections.

The professors Michael Muma and Daniel Palomar supervised the work. The PhD project has been supported by the DFG project REFOCUS (2019-2022) and by the LOEWE Center emergenCITY. Jasin Machkour will remain at TU Darmstadt and the LOEWE Center emergenCITY as a post-doc.

Acknowledgments

We would like to thank the committee chairman, Ralf Steinmetz, as well as professors Christoph Hoog Antink and Grace Li Zhang for their efforts. We would also like to thank Daniel Palomar for making the long journey from Hong Kong to Darmstadt.

Author:

Michael Muma, Principal Investigator at emergenCITY and Head of the Robust Data Science Group at the Institute of Communications Engineering at TU Darmstadt.