Menü
Head of the Robust Data Science Group, TU Darmstadt.
Interests:
Michael Muma is Head of the Robust Data Science Group at Technische Universität Darmstadt. His research is on new robust data science theory and methods with applications in signal processing and machine learning for biomedicine and engineering.
He is the Principal Investigator of the ERC Starting Grant ScReeningData. He is also PI within the LOEWE center emergenCITY and the BMBF Cluster4Future curATime. He is a lecturer for Robust and Biomedical Signal Processing and the supervisor or co-supervisor of 11 PhD students (6 completed). From 2017-2022, he was Independent Junior Research Group Leader (Athene Young Investigator) at the Signal Processing Group of Technische Universität Darmstadt, where he received his PhD in 2014.
He received the 2021 Early Career Award of the European Association For Signal Processing (EURASIP) for his contributions to robust signal processing and statistical learning. Since 2022, he is Chair of the Technical Area Committee on Theoretical and Methodological Trends in Signal Processing of EURASIP. Since 2019, he serves as Associate Editor for the IEEE Transactions on Signal Processing and since 2022, he serves as Associate Editor for the IEEE Open Journal on Signal Processing. In 2017, together with his co-authors, he received the 2017 IEEE Signal Processing Magazine Best Paper Award for the article Robust Estimation in Signal Processing: A tutorial-style treatment of fundamental concepts. He was the supervisor of the 2020 IEEE Radar Conference Student Best Paper Award winner and the student team who achieved first place in the international IEEE Signal Processing Cup 2015.
He was the Lead Guest Editor of the 2019 Elsevier Signal Processing Special Issue on Statistical Signal Processing Solutions and Advances for Data Science: Complex, Dynamic and Large-scale Settings. He is active in organizing scientific events, such as, the 2021 Statistical Learning for Signal and Image Processing (SLSIP) Workshop and the Joint IEEE SPS and EURASIP Summer School on Robust Signal Processing. He is also co-author of the book Robust Statistics for Signal Processing that was published in 2018 with Cambridge University Press. At the 45th IEEE ICASSP 2020, he was tutorial speaker on Robust Data Science: Modern Tools for Detection, Clustering and Cluster Enumeration.
Michael Muma is Head of the Robust Data Science Group at Technische Universität Darmstadt. His research is on new robust data science theory and methods with applications in signal processing and machine learning for biomedicine and engineering.
He is the Principal Investigator of the ERC Starting Grant ScReeningData. He is also PI within the LOEWE center emergenCITY and the BMBF Cluster4Future curATime. He is a lecturer for Robust and Biomedical Signal Processing and the supervisor or co-supervisor of 11 PhD students (5 completed). From 2017-2022, he was Independent Junior Research Group Leader (Athene Young Investigator) at the Signal Processing Group of Technische Universität Darmstadt, where he received his PhD in 2014.
He received the 2021 Early Career Award of the European Association For Signal Processing (EURASIP) for his contributions to robust signal processing and statistical learning. Since 2022, he is Chair of the Technical Area Committee on Theoretical and Methodological Trends in Signal Processing of EURASIP. Since 2019, he serves as Associate Editor for the IEEE Transactions on Signal Processing and since 2022, he serves as Associate Editor for the IEEE Open Journal on Signal Processing. In 2017, together with his co-authors, he received the 2017 IEEE Signal Processing Magazine Best Paper Award for the article Robust Estimation in Signal Processing: A tutorial-style treatment of fundamental concepts. He was the supervisor of the 2020 IEEE Radar Conference Student Best Paper Award winner and the student team who achieved first place in the international IEEE Signal Processing Cup 2015.
He was the Lead Guest Editor of the 2019 Elsevier Signal Processing Special Issue on Statistical Signal Processing Solutions and Advances for Data Science: Complex, Dynamic and Large-scale Settings. He is active in organizing scientific events, such as, the 2021 Statistical Learning for Signal and Image Processing (SLSIP) Workshop and the Joint IEEE SPS and EURASIP Summer School on Robust Signal Processing. He is also co-author of the book Robust Statistics for Signal Processing that was published in 2018 with Cambridge University Press. At the 45th IEEE ICASSP 2020, he was tutorial speaker on Robust Data Science: Modern Tools for Detection, Clustering and Cluster Enumeration.
He received the 2021 Early Career Award of the European Association For Signal Processing (EURASIP) for his contributions to robust signal processing and statistical learning. Since 2022, he is Chair of the Technical Area Committee on Theoretical and Methodological Trends in Signal Processing of EURASIP. Since 2019, he serves as Associate Editor for the IEEE Transactions on Signal Processing. In 2017, together with his co-authors, he received the 2017 IEEE Signal Processing Magazine Best Paper Award for the article Robust Estimation in Signal Processing: A tutorial-style treatment of fundamental concepts. He was the supervisor of the 2020 IEEE Radar Conference Student Best Paper Award winner and the student team who achieved first place in the international IEEE Signal Processing Cup 2015.
He was the Lead Guest Editor of the 2019 Elsevier Signal Processing Special Issue on Statistical Signal Processing Solutions and Advances for Data Science: Complex, Dynamic and Large-scale Settings. He is active in organizing scientific events, such as, the 2021 Statistical Learning for Signal and Image Processing (SLSIP) Workshop and the Joint IEEE SPS and EURASIP Summer School on Robust Signal Processing. He is also co-author of the book Robust Statistics for Signal Processing that was published in 2018 with Cambridge University Press. At the 45th IEEE ICASSP 2020, he was tutorial speaker on Robust Data Science: Modern Tools for Detection, Clustering and Cluster Enumeration.
Terminating-Random Experiments (T-Rex) methods provide a new scalable learning framework to ensure reproducibility in high-dimensional data. Reproducible discoveries are achieved by controlling the false-discovery-rate (FDR), while simultaneously maximizing the true-positive-rate (TPR, “Power”). The underlying computer-assisted statistical learning can be compared to a placebo-controlled study in drug development: (i) randomized controlled experiments are systematically computed on the computer and mathematically modeled (ii) discoveries are only declared as reproducible if they are sufficiently successful against computer-generated placebo markers („dummies“). A speed advantage of multiple orders of magnitude over existing methods comes from the fact that learning is stopped early (termination) when dummies are selected.
Reliable and robust information extraction and processing in complex data sets is essential in today’s data science. In practice, one often encounters corrupted measurements, impulsive noise, artifacts and outliers. Robust methods systematically deal with such uncertainties and corruptions within the mathematical framework of robust statistics. Therefore, in contrast to classical procedures they are not significantly affected. Developing robust methods has been the core-focus of Michael Muma’s research. He regularly publishes his results in top journals and conferences. He is the EURASIP Early Career Awardee 2021 for his contributions to robust signal processing and statistical learning, and IEEE Signal Processing Magazine Best Paper Awardee for the article “Robust estimation in signal processing: A tutorial-style treatment of fundamental concepts.” He is also co-author of the textbook “Robust statistics for signal processing” published 2018 with Cambridge University Press.
Michael Muma has more than 10 years of experience in biomedical engineering, leading projects in interdisciplinary teams with medical practitioners, industry and academia. Applications include but are not limited to cardiovascular signals from various sensors, such as ECG and PPG, eye research, intracranial pressure, psychology, radar-based gait analysis, hearing aids, and health-monitoring using wearables.