Jasin Machkour

Current Members

Dr.-Ing. Jasin Machkour

Member of the Robust Data Science Group at the Institute of Telecommunications, TU Darmstadt.

Contact

jasin.machkour@tu-...

work +49 6151 16-22531

Work S3|06 29
Merckstr. 25
64283 Darmstadt

Links

Office Hours: Upon request

Bio

Jasin received the Dr.-Ing. degree (equivalent to Ph.D.), the M.Sc. degree in Electrical Engineering and Information Technology with a major in Communication and Sensor Technology, and the B.Sc. degree in Industrial Engineering from Technische Universität Darmstadt (TU Darmstadt) in 2024, 2019, and 2016, respectively.

He is a founding member of the EURASIP Student Committee and represents the area Theoretical and Methodological Trends in Signal Processing (TMTSP).

From September 2019 until February 2020 and from May 2023 until August 2023, he visited Prof. Daniel P. Palomar at The Hong Kong University of Science and Technology. From May 2022 until July 2022, he visited Prof. Frédéric Pascal at Université Paris-Saclay (CentraleSupélec). In 2022, he completed the Statistical Genomics Summer Course at University of Oxford.

He spent one year of his master’s studies (2016/2017) at the Department of Electrical and Computer Engineering of the University of Illinois at Urbana-Champaign (USA).

Research

Jasin’s research interests lie in the broad field that is often called statistical learning, machine learning, data science, or statistical signal processing. He is particularly interested in developing methods/algorithms for large-scale and high-dimensional learning tasks with provable statistical properties.

The following topics are of great interest to him:

Sparse regression in high-dimensional settings
False discovery rate (FDR) control in variable/feature selection
Methods for genome-wide association studies (GWAS)
Parallel computing on high-performance computing (HPC) clusters.

Software

TRexSelector: T-Rex Selector: High-Dimensional Variable Selection & FDR Control [CRAN] [GitHub] [Paper]

tlars: The T-LARS Algorithm: Early-Terminated Forward Variable Selection [CRAN] [GitHub] [Paper]

Fabian Scheidt	Robust and Computationally Efficient Statistical Learning in High-Dimensional Data: The T-Rex Selector	Master's Thesis	03/2023
Marie Flohr	Reproducibility Analysis of Discoveries in Genome-Wide Association Studies Using Variable Selection Methods for High-Dimensional Data	Bachelor's Thesis	12/2022
Simon Tien (co-supervision with Michael Muma)	Development of High-Dimensional Learning Methods for Genome-Wide Association Studies	Master's Thesis	06/2022
Lisa Dawel (co-supervision with Michael Muma)	Development and Analysis of Sparse High-Dimensional Ensemble Learning Methods	Master's Thesis	12/2021
Pertami Kunz	Robust Variable Selection and False Discovery Rate Control for High-Dimensional Regression Models Using Knockoffs	Master's Thesis	02/2021
Qiang Zhao	Robust Error Control for High-Dimensional Variable Selection based on the Stability Selection Method	Master's Thesis	01/2021
Group of three Master's students	Python vs. R: Benchmarking the computational performance of the Terminating Random Experiments (T-Rex) Method	RSP Seminar	07/2022
Group of four Master's students	HIV Dataset Case-Study C – Same Features but Different Responses	RSP Seminar	02/2022
Individual project (Master's student)	Integrating Robust Loss Functions Into Deep Learning	RSP Seminar	02/2021
Individual project (Master's student)	Robust Reconstruction of Biomedical Images in the Presence of Outliers	RSP Seminar	02/2021
Individual project (Master's student)	HIV Dataset Case-Study B – What Types of Outliers are More Severe?	RSP Seminar	02/2021
Group of two Master's students	Dependent Data Bootstrap	ATISSP Seminar	02/2021
Group of two Master's students	Hypotheses Testing With The Bootstrap and Signal Detection	ATISSP Seminar	02/2021
Group of two Master's students	Applications of The Bootstrap	ATISSP Seminar	02/2021