Mr Prof. Dr. Peter Zaspel
Biography
| since July 2023 | W2 Professorship in Software for Data-Intensive Applications, University of Wuppertal, Wuppertal, Germany |
| March 2022 – June 2023 | Assistant Professor of Computer Science (Machine Learning), Jacobs University Bremen gGmbH, Bremen, Germany |
| Aug. 2019 – Feb. 2022 | Interim Professor of Computer Science (Machine Learning), Jacobs University Bremen gGmbH, Bremen, Germany |
| 2017 – 2019 | Postdoc, Department of Mathematics and Computer Science at the University of Basel (DMI, topical area: mathematics), Basel, Switzerland. |
| 2015 – 2017 | Postdoc, Heidelberg Institute for Theoretical Studies: HITS gGmbH, Heidelberg, Germany. |
| 2015 – 2017 | Postdoc (associated), Interdisziplinary Center for Scientific Computing (IWR, University of Heidelberg), Heidelberg, Germany. |
| 2009 – 2015 | Research assistant, Institute for Numerical Simulation (INS, University of Bonn), Bonn, Germany. |
Education
| 2019 – 2021 | Habilitation in mathematics, University of Basel, Basel, Switzerland. |
| 2009 – 2015 | PhD student in applied mathematics, University of Bonn, Bonn, Germany. |
| 2004 – 2009 | Diplom student in computer science, University of Bonn, Bonn, Germany. |
Ongoing research projects
- DFG project “MFML-OT – Multi-Fidelity Machine Learning for Orthogonal Translation” (duration 2027-2029)
Research project within the DFG SPP 2546 “Artificial Intelligence for Protein Function” (joint project with Mehdi D. Davari, Leibniz Institute of Plant Biochemistry, Halle, and Michael A. Nash, University of Basel / ETH Zurich) - DFG project “ICEBAY – Temperature reconstruction combining boreholes thermometry and ice-cores with Bayesian hierarchical modelling – Global warming vs. natural variability in DML, Antarctica” (duration: 2026-2028)
Research project within the DFG SPP 1158 “Antarctic Research” (joint project with Thomas Laepple, Alfred Wegener Institut, Potsdam) - DFG project “Multi-fidelity, active learning strategies for exciton transfer in cryptophyte antenna complexes” (duration 2026-2029)
Research project within the second phase of the DFG SPP 2363 “Utilization and Development of Machine Learning for Molecular Applications – Molecular Machine Learning” (joint project with Ulrich Kleinekathöfer, Physics, Constructor University Bremen) - DFG subproject “Data-driven surrogate modelling for differential-algebraic port-Hamiltonian systems” (C03) (duration: 2025-2028)
Subproject of the CRC 1701 “Port-Hamiltonian Systems” (co-PI together with Michael Günther) - DFG subproject “Benchmark platform for port-Hamiltonian systems” (S) (duration: 2025-2028)
Subproject of the CRC 1701 “Port-Hamiltonian Systems” (co-PI together with Hanno Gottschalk)
Research interests
- Data-efficient and multifidelity machine learning
- Uncertainty quantification and Bayesian methods
- Scalable kernel methods and algorithms
- Scientific applications driving methodological development
Publications
Publication list
Invited talks
Augmenting the explanatory power of predictions by uncertainty quantification,
2nd Workshop on Embedded Machine Learning – WEML2018, Heidelberg University, Nov 8, 2018.
Meshfree and multi-index approximations for parametric real-world problems,
Seminar on Uncertainty Quantification, RWTH Aachen, Aachen, Germany, August 29, 2018.
Optimal-complexity kernel-based stochastic collocation with application in fluid mechanics,
Seminar of the “Mathematics in Computational Science and Engineering” group, on invite by Prof. Dr. Fabio Nobile, EPFL, Lausanne, Switzerland, October 24, 2017.
Scalable solvers for meshless methods on many-core clusters,
QUIET 2017 – Quantification of Uncertainty: Improving Efficiency and Technology, SISSA, International School for Advanced Studies, Trieste, Italy, July 18-21, 2017.
H-matrices on many-core hardware with applications in parametric PDE’s,
Colloquium of the Faculty of Engineering, on invite by Prof. Dr. Steffen Börm, University of Kiel, December 9, 2016.
Algorithmic patterns for hierarchical matrices on many-core processors,
Seminar in Numerical Analysis, on invite by Prof. Dr. Helmut Harbrecht, University of Basel, September 18, 2016.
