About me
- Postdoctoral Researcher at TimeXAI Research Group at the Centre of Competence for Information Technology at Technische Hochschule Mittelhessen - University of Applied Sciences in Friedberg, Hesse
- Lecturer at THM StudiumPlus Dual Study Programme, Modules: Machine Learning, Predictive Analytics, Computer Architectures and Operating Systems, Compiler Design
Main Research Interest: Enabling Trustworthy and Safe Use of AI in Medical and Industrial Practice
I am a postdoctoral researcher at Technische Hochschule Mittelhessen in Friedberg, Germany. I earned my Doctor of Engineering from the Graduate Centre for Engineering Sciences at the Research Campus of Central Hessen in Cooperation with the Philipps-University of Marburg. With an industry-earned background in data analytics, my research focuses on applied artificial intelligence in medicine, particularly in the field of explainable AI (XAI) and time series analysis. My scientific contributions include the development of SIGN, a novel XAI method, and its evaluation against established clinical guideline patterns. I have also published work on the detection of myocardial scar and structural changes in myocardium in general, using electrocardiogram (ECG) data. My primary interest lies in making AI applicable in practice through the development of interpretable and transparent models.
N. Gumpfer et al. (2023). SIGNed Explanations: Unveiling Relevant Features by Reducing Bias. Information Fusion 99, p. 101883, DOI:10.1016/j.inffus.2023.101883
N. Gumpfer et al. (2024). Towards Trustworthy AI in Cardiology. In: AIME ’24, Salt Lake City, UT, USA, July 9 - 12, 2024. Vol. 14845. LNCS, pp. 350–361, DOI:10.1007/978-3-031-66535-6_36
N. Gumpfer et al. (2023). SIGNed Explanations: Unveiling Relevant Features by Reducing Bias. Information Fusion 99, p. 101883, DOI:10.1016/j.inffus.2023.101883