PhD position Computational aspects of Bayesian inverse problems with non-Gaussian priors

Research / Academic

In applications such as image processing and materials science, we wish to understand a complex system from noisy observations. This is the challenging setting of  inverse problems. The PhD project concerns employing a Bayesian approach to such inverse problems with general, non-Gaussian priors, with the aim of preserving structure available in the information. Examples of non-Gaussian priors are Besov priors which
allow function discontinuities (similar in spirit to the TV regularization used in image processing). Additional challenges may be how to deal with situations where the likelihood is not differentiable or the likelihood computation is of black-box type and/or has no gradient information available.
One of the main challenges you will address is how to design suitable (Monte Carlo) algorithms for carrying out computations for such models. After first performing an exploration of the existing methods and literature available on the topic, the project may focus more on theoretical or applied aspects, depending also on the interests of the student. 
This project is part of the Delft AI Lab ‘SLIMM-Lab’, a collaboration between Hanne Kekkonen and Joris Bierkens (Delft Institute of Applied Mathematics) and Iuri Rocha and Frans van der Meer (Civil Engineering and Geosciences). The goal of SLIMM Lab is to employ machine learning approaches towards efficient computation in materials science. The current PhD position focusses on some of the mathematical underpinnings
of these approaches.


  • MSc degree in mathematics with a specialization or strong affinity with one of the following: statistics, probability, machine learning analysis or numerical analysis.
  • Affinity with coding numerical methods, e.g. in Python, C++, or Julia.
  • Interest or experience in teaching and guiding students combined with a strong scientific drive.
  • Ability to work in a multidisciplinary and diverse team.
  • Good English proficiency, both verbally and in writing. 

Doing a PhD at TU Delft requires English proficiency at a certain level to ensure that the candidate is able to communicate and interact well, participate in English-taught Doctoral Education courses, and write scientific articles and a final thesis. For more details please check the Graduate Schools Admission Requirements.

Salary Benefits:

TU Delft offers DAI-Lab PhD-candidates a 5-year contract (as opposed to the normal 4-years), with an official go/no go progress assessment after one year. Salary and benefits are in accordance with the Collective Labour Agreement for Dutch Universities, increasing from € 2541 per month in the first year to € 3413 in the fifth year. As a PhD candidate you will be enrolled in the TU Delft Graduate School. The TU Delft Graduate School provides an inspiring research environment with an excellent team of supervisors, academic staff and a mentor. The Doctoral Education Programme is aimed at developing your transferable, discipline-related and research skills.
The TU Delft offers a customisable compensation package, discounts on health insurance and sport memberships, and a monthly work costs contribution. Flexible work schedules can be arranged. For international applicants we offer the Coming to Delft Service and Partner Career Advice to assist you with your relocation.

Work Hours:

36 - 40 hours per week


Mekelweg 2