PhD Candidate Causal inference & Machine Learning

About your role

To truly support medical decision making, prediction algorithms need to take the causal effects of interventions into account. There is an increasing interest in so-called `prediction under intervention’ methods, that can provide patient prognoses under alternative treatment scenarios (e.g., ‘what if this patient initiates treatment A’). However, regular model assessment techniques such as cross-validated AUC or calibration cannot be used in this context since ground truth causal quantities are never observed. To solve this problem, you will develop novel performance criteria to evaluate causal prediction algorithms, with a focus on time-to-event outcomes.

To test feasibility, you will apply these methods to high-stakes medical prediction algorithm in transplantation and transfusion research. To support the uptake of your methods, you will create user-friendly statistical software, such as an R package. You will share your findings through academic publications, conference presentations, and collaborations with clinical researchers. Additionally, you will actively contribute to the institute's academic environment by participating in teaching, workshops, and seminars; teaching activities account for approximately 10% of your appointment

You will be part of an interdisciplinary and international consortium of researchers from biostatistics, computer science, mathematics and epidemiology (the Safe Causal Inference Consortium). A key aim of our consortium is to take an interdisciplinary approach and connect results from different quantitative fields.

Your main tasks will be to:

• Develop new performance criteria for ‘prediction under interventions’ and develop computational algorithms to estimate these performance criteria.
• Derive their statistical properties, through mathematical derivations and simulation studies.
• Apply the algorithms to large medical datasets and interpret findings.

About your workplace

You will work within the Department of Biomedical Data Sciences at LUMC, where where advanced statistical methods, data science, and data management are developed and applied to biomedical research. The department plays a key role in strengthening research quality and supporting medical decision-making through data-driven insights.

 Within this department, you will be part of the Medical Statistics section that consists of a large group of biostatisticians. You will join a close-knit and highly skilled team of PhD candidates and senior researchers working at the intersection of prediction modeling and causal inference. The working environment is open, intellectually stimulating, and collegial, offering ample room for in-depth discussion, collaboration, and personal development.

This project is part of the ‘Safe Causal Inference’ consortium (8 PhD candidates in total). You will work closely together with causal inference researchers from Delft University of Technology, Erasmus Medical Center and VU Amsterdam, combining expertise from computer science, mathematics, biostatistics and epidemiology. In addition, you will benefit from access to an international network of experts in causal inference.

You will be supervised by an interdisciplinary team, covering expertise in biostatistics (Dr Nan van Geloven), computer science (Dr Jesse Krijthe) and epidemiology (Prof Saskia le Cessie).

About you

As a PhD candidate in causal inference & machine learning you are fascinated by developments in causal AI, and motivated to use your quantitative skills to advance healthcare. You are curious to find out how things work and you want to get to the bottom of things. You enjoy explaining your findings to others, including researchers with a different scientific background. You are at ease with programming (in eg R or Python).

In addition, you bring:

• A master’s degree in (Bio)Statistics, Applied Mathematics, Data Science, Econometrics or a related field, with a strong academic track record, for instance an excellent thesis (≥8.0/10 on the Dutch scale or ECTS Grade A/B).
• An interest in both studying the theoretical properties of estimators and applying statistical methods to medical datasets.
• Experience with programming, preferably in R and/or Python.
• Strong written and oral communication skills in English (C1 level or higher).
• Proficiency in Dutch is considered an advantage.

Our offer

Driven by health; that's our mission. This applies not only to our patients but also to our employees. We strive to provide a safe, inclusive and equitable work environment, with a focus on talent development and connected leadership. In order to be able to continue to learn and develop, we offer internal and external training. You are also entitled to an end-of-year bonus (8,3%), holiday allowance, sports budget, bicycle scheme, home office allowance and an excellent commuting allowance. Furthermore, as an employee of LUMC, you are also affiliated with the ABP pension fund. This means that 70% of your pension premium is paid by LUMC, leaving you with a higher net salary. Nice, right?

See all employment conditions of the LUMC.

What do we represent?

As an academic medical center, LUMC strives to have a workforce that is a good reflection of society. We are committed to the highest quality in health care, education and (international) research, in which diverse perspectives are essential. Sustainability is also a high priority for us: we are committed to a healthy future, not only for people, but also for the planet. In line with these values, we aim to be a creative and inspiring place to work, where everyone feels at home, safe and valued. It's all about who you are and what you bring to contribute, regardless of your background. Together, we work toward a sustainable and inclusive future, both for the people around us and for the world we live in. Together, we are LUMC.  

More information

• As a PhD-candidate, your salary will be between €3,108 and €3,939 gross per month (scale Pro, CAO UMC). These figures are based on a full-time position.
• This position is for the duration of 4 years
• Please upload your CV, motivation letter (max 1 A4) as well as your MSc transcripts
• The procedure involves two to three interviews, including one where we ask you to present some of your earlier research work (e.g. you master’s thesis). The recruitment committee consists of Dr Nan van Geloven, Dr Jesse Krijthe and Prof Saskia le Cessie.
• Applications from employment/recruitment agencies will not be considered.