News

emergenCITY scientist develops T-Rex selector, an efficient framework that enables fast variable selection in large, high-dimensional data sets.

Finding missing people under rubble after an earthquake is a major challenge, even for rescue robots. The situation is similar in medicine: finding genes that are responsible for specific diseases is complicated. Sometimes it’s like looking for a needle in a haystack. For applications such as these, emergenCITY scientist Jasin Machkour has developed an efficient method, the T-Rex selector. A paper on the T-Rex Selector has now been published in Signal Processing, one of the most renowned journals for signal processing, and another paper has already been accepted for publication.

On Wednesday, November 20, 2024, the “Athena Prizes for Good Teaching” were awarded.

Designing Human-AI Collaborative Workflows for Advanced Image Analysis in Physics

Project Overview: This Masters Project analyzes about one million diffraction images from the Linac Coherent Light Source (LCLS) at the Stanford Linear Accelerator Center (SLAC).

On August 23, 2024, Mr. Jasin Machkour defended his doctoral thesis titled “Development of Fast Machine Learning Algorithms for False Discovery Rate Control in Large-Scale High-Dimensional Data”.

T-Rex based high-dimensional Variable Selection with False Discovery Rate Control.

As part of the curATime project, supported by the Cluster4Future initiative of the Federal Ministry of Education and Research, a workshop of the cluster project entitled “T-Rex based high-dimensional Variable Selection with False Discovery Rate Control” took place. It was organized by the research groups for Robust Data Science and Signal Processing at TU Darmstadt in collaboration with the Center for Thrombosis and Hemostasis (CTH) of the University Medical Center Mainz at the University Medical Center Mainz.

Conduct your Masters Project under joint supervision of the Robust Data Science Group and Biophotonics – Biomedical Engineering Group to enhance your skills in machine learning, signal processing, laser spectroscopy, and biomedical engineering.

Statistical Learning for Signal and Image Processing (SLSIP) Workshop in Porquerolles, South of France.

“Physiological and Emotional-Annotated Time Series Analysis for Affective Computing and Biomedical Applications“.

You are an expert in C++ and interested in biomedical applications? We are offering a Student Assistant (HiWi) position to engage in cutting-edge research that concerns computationally efficient C++ implementations for the analysis of large-scale biomedical data. While small data sets can be processed with python or R, applying advanced statistical machine learning algorithms to large-scale biobanks requires backends or complete implementations in C++. Efficiency is required with regards to algorithmic operations, but also efficient memory handling/memory mapping and online-processing.

Fabian Scheidt received a short-term financial grant to participate at the IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP) in Costa Rica.

Thanks to the support of the emergenCity, Fabian Scheidt, a PhD. Student in Electrical Engineering, Signal Processing and Robust Data Science, from TU Darmstadt’s Robust Data Science Group of Prof. Michael Muma 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.

State funds research network that aims to make cities more resilient to crises and disasters

Wiesbaden/Darmstadt. The Darmstadt LOEWE center emergenCITY will continue to receive funding from the state of Hesse and will receive a total of around five million euros in project funding for 2024. This was decided by the committees of the state's LOEWE research funding program.

Jasin Machkour visited Prof. Palomar at The Hong Kong University of Science and Technology (HKUST) to work on the development of false discovery rate (FDR) controlling methods for dependent variables.

From 28.09.2023 to 01.10.2023, members of the Robust Data Science Group of TU Darmstadt visited the CERN research center in Geneva. The excursion was organized by the Signal Processing Group. Many thanks to our colleagues!

Masters Lecture Robust Data Science With Biomedical Applications Begins October 30th

The lecture covers fundamental topics and recent developments in robust data science. Unlike classical statistical learning and signal processing, which relies strongly on the normal (Gaussian) distribution, robust methods can tolerate impulsive noise, outliers and artifacts that are frequently encountered in biomedical applications.