master thesis projects

Physics-Guided Machine Learning - A Novel Neural Network-Based Approach for Inverse Problems

Co-Supervisor(s): Jing Sun (TU Delft, Intelligent Systems)

Operator learning for periodic systems with partially hidden physics

Co-Supervisor(s): Birupaksha Pal (Bosch, India)

Stabilization Methods for Physics-Informed Neural Networks

Co-Supervisor(s): Franziska Griese, Philipp Knechtges (DLR)

Machine Learning-Accelerated Solvers for Computational Fluid Dynamics Simulations

Co-Supervisor(s): Merten Stender (TU Berlin)

Decomposing Graph Neural Networks

Co-Supervisor(s): Alena Kopaničáková (University of Toulouse)

Domain Decomposition Techniques for the Helmholtz Equation

Co-Supervisor(s): Vandana Dwarka (TU Delft, Numerical Analysis)

Numerical techniques for efficiently solving a nonlinear model for salt intrusion in rivers

Co-Supervisor(s): Yoeri Dijkstra (TU Delft, Mathematical Physics)

Improving Nonlinear Solver Convergence Using Machine Learning

Co-Supervisor(s): Tycho van Noorden (COMSOL)

Overlapping Schwarz Domain Decomposition Methods for Implicit Ocean Models

Co-Supervisor(s): Jonas Thies (TU Delft, Numerical Analysis)

Parallel Multiplicative One-Level Schwarz Preconditioners With FROSch and Trilinos

Spectral Neural Operators

Student: Hangyu Xia

Investigation of the convergence of deep operator networks

Student: Johannes Taraz

Machine learning for post-storm profile predictions

Co-Supervisor(s): José Antonio Álvarez Antolínez (TU Delft, Hydraulic Engineering), Renske Gelderloos (TU Delft, Environmental Fluid Mechanics), Panos Athanasiou, Kees Nederhoff (Deltares)

Student: Derby Oudenes

Geometry Learning for Complex Shaped Cells

Co-Supervisor(s): Qiyao Peng (Vrije Universiteit Amsterdam)

Student: Ikhlasse Ouadah

Data-driven methods for machine-induced metrology errors

Co-Supervisor(s): Sjoerd Wiarda (ASML)

Student: Michael Aldorf

Error Estimates for Finite Element Simulations Using Neural Networks

Co-Supervisor(s): Deepesh Toshniwal (TU Delft, Numerical Analysis)

Student: Yijun Wu

Block Preconditioners for Monolithic Solvers of Very Large Floating Structures

Co-Supervisor(s): Oriol Colomés (TU Delft, Offshore Engineering)

Student: Shreyas Prashanth

Creating a machine learning based outlier removal algorithm that incorporates a priori knowledge of the physics

Co-Supervisor(s): Lucas Bekker (ASML)

Student: Thomas Kamminga

Separating longtime behavior and learning of mechanisms

Co-Supervisor(s): Cordula Reisch (TU Braunschweig)

Student: Tirtho Sarathi Saha

Machine learning for post-storm profile predictions

Co-Supervisor(s): José Antonio Álvarez Antolínez, Ad Reniers (TU Delft, Hydraulic Engineering), Panos Athanasiou, and Robert McCall (Deltares)

Student: Koen van Asselt

Estimating congestion and traffic patterns when planning road work

Co-Supervisor(s): Leonoor Portengen, Henk van Haaster (CGI)

Student: Vosse Meijssen

Surrogate models for the characterization of hydrodynamic loads on perforated monopiles

Co-Supervisor(s): Oriol Colomés (TU Delft, Offshore Engineering)

Student: Ruben Dekeyser

Domain Decomposition for Machine Learning Based Medical Imaging

Co-Supervisor(s): Eric Cyr (Sandia National Laboratories, USA)

Student: Corné Verburg

Predicting the Optimal Solver Settings with Machine Learning in COMSOL CFD Module

Co-Supervisor(s): Tycho van Noorden (COMSOL)

Student: Anouk Zandbergen

AI-driven turbulence modeling of two-phase flows in nuclear reactors

Co-Supervisor(s): Edo Frederix (NRG), Deepesh Toshniwal (TU Delft, Numerical Analysis)

Student: Gonzalo Bonilla Moreno

Convolutional Neural Networks for Time-Dependent Fluid Flow

Student: Sylle Hoogeveen

Machine Learning for Turbulent Flows

Co-Supervisor(s): Werner Lazeroms (Sioux)

Student: Elske van Leeuwen

Enhancing Sea Ice Dynamics in the Climate System by Machine Learning

Co-Supervisor(s): Carolin Mehlmann (Otto-von-Guericke University, Max-Planck-Institute of Meteorology)

Student: Brendan Analikwu

Domain Decomposition Techniques for the Helmholtz Equation

Co-Supervisor(s): Vandana Dwarka (TU Delft, Numerical Analysis)

Student: Erik Sieburgh

Surrogate models for the characterization of hydrodynamic loads on perforated monopiles

Co-Supervisor(s): Oriol Colomés (TU Delft, Offshore Engineering)

Student: Quinten Star

The Application of Neural Networks to Predict Skin Evolution After Burn Trauma

Co-Supervisor(s): Fred Vermolen (University of Hasselt)

Student: Selma Husanović

Reduced Order Models for Fluid Flow With Generative Adversarial Networks (GANs)

Co-Supervisor(s): Kees Vuik (TU Delft, Numerical Analysis)

Student: Mirko Kemna