KU Leuven's NUMA (Numerical Analysis and Applied Mathematics) research division, part of the Department of Computer Science, is offering a fully funded PhD position focused on the development of multilevel interacting particle methods for Bayesian inversion. NUMA is a vibrant research environment with 12 permanent staff and around 60 PhD and postdoctoral researchers, dedicated to advancing numerical algorithms and software for large-scale scientific and engineering problems. The project addresses the challenge of calibrating mathematical models—particularly those governed by partial differential equations (PDEs) with high-dimensional or infinite-dimensional parameters—using measurement data. Such inverse problems are computationally intensive, especially when high-resolution data is involved. The research will center on designing and analyzing efficient Bayesian computational methods for these ill-posed inverse problems, with a special emphasis on improving the computational efficiency of interacting particle methods within a multilevel framework that accounts for model error. Applications of the developed methods will be drawn from structural mechanics, benefiting from NUMA's longstanding collaboration with the Structural Mechanics group in the Department of Civil Engineering. However, the research is fundamentally methodological and not tied to a single application domain. The successful candidate will join an international team of numerical analysts and have the opportunity to develop expertise in state-of-the-art simulation and computational techniques. Eligibility: Applicants must hold a master's degree in Mathematical Engineering, (Applied) Mathematics, or an equivalent field. A strong background in numerical methods for differential equations, stochastic process simulation, and/or optimization is required, as is experience in scientific software programming. Excellent English proficiency and strong communication skills are essential. Funding: The position is fully funded for four years, with a competitive salary and access to a supportive, collaborative research environment. The project offers excellent opportunities for professional development in both academic and industrial research contexts. Application Process: Interested candidates should submit a letter of motivation (including a statement of skills and research interests), a curriculum vitae, copies of university diplomas and transcripts, and the names and contact information of 1-2 references. Applications are reviewed on a rolling basis until the position is filled, with a formal deadline of April 30, 2026. For further information, contact Prof. dr. ir. Giovanni Samaey at Giovanni.Samaey@kuleuven.be. KU Leuven is committed to diversity, inclusion, and equal opportunity, fostering an open and respectful academic environment. For questions about accessibility or support, applicants are encouraged to reach out via the provided contact email.

NUMA, the Numerical Analysis and Applied Mathematics section within the Department of Computer Science at KU Leuven, is offering a postdoctoral position focused on interacting particle methods for Bayesian inversion with model error. NUMA is a vibrant research group comprising 12 permanent staff and around 60 PhD and postdoctoral researchers, dedicated to developing advanced numerical algorithms and software for large-scale scientific and engineering problems. This project addresses the challenge of reliably using simulation-generated predictions in science and engineering, particularly when mathematical models—often partial differential equations (PDEs) with infinite-dimensional parameters—must be calibrated to measurement data. The research centers on computationally intensive inverse problems, where parameters and states are inferred from large, high-resolution datasets. The team aims to advance Bayesian computational methods for these ill-posed inverse problems, focusing on both increasing their validity and reducing computational cost. The specific goal of this postdoc is to enhance the validity of interacting particle methods for Bayesian inversion by explicitly including model error in the likelihood evaluation. The model problem will involve inferring parameters in phenomenological models for cardiac electrophysiology, leveraging a long-standing collaboration with the Department of Cardiovascular Imaging and Dynamics. The work will be applied in the context of cardiac excitation simulations, but the methods developed are intended to be generic and broadly applicable. Candidates should hold a PhD in Mathematical Engineering or Applied Mathematics (or equivalent), with a strong background in numerical methods for differential equations, simulation of stochastic processes, and/or optimization. Experience with sampling methods for Bayesian inversion and cardiac electrophysiological modeling is highly valued. Proficiency in programming scientific software and excellent English communication skills are required. The position offers a high-level international research environment, a supportive and collaborative team, and opportunities to develop expertise in state-of-the-art simulation methods. Funding is secured for two years, with the possibility of a third year depending on progress and available resources. The salary is competitive. Applications should include a letter of motivation, CV, diploma copies and transcripts, and contact information for 1-2 references. The formal deadline is February 28, 2026, but applications will be reviewed as received and the position will close once a suitable candidate is found. For further information, contact Prof. dr. ir. Giovanni Samaey at Giovanni.Samaey@kuleuven.be. KU Leuven is committed to diversity, inclusion, and equal opportunity, fostering an environment of open dialogue and respect for all backgrounds.