Profile photo
Roel Hulsman
PhD candidate in causal ML

I am a second-year PhD candidate at AMLab in Amsterdam, supervised by Sara Magliacane and Herke van Hoof. My PhD is funded by Adyen, where I spent a minor portion of my time (20%). I study causal dynamics in non-stationary settings, including causal discovery, causal representation learning, performative prediction and multi-armed bandits, mostly focusing on time series. My latest work is on the identifiability of Markov switching models (MSMs), accepted at ICML'26.

I graduated with distinction from the University of Oxford with a master's in statistics, where I was supervised by Rob Cornish and Arnaud Doucet. I also obtained a BSc in econometrics and operations research and a BA in philosophy of the social sciences from the University of Groningen, both cum laude. Before starting my PhD, I spent a short period at ASML as a data analyst for business intelligence, and at the Joint Research Centre (JRC) of the European Commission in Italy on AI for healthcare.

For the upcoming academic year (2026-2027), I will supervise thesis projects for the master in AI at the University of Amsterdam. Unfortunately I am not able to supervise students in other programs or universities. If you are a master student in AI at UvA and you are interested to work with me, feel free to reach out!

Resume

Publications

ICML'26
Roel Hulsman, Carles Balsells-Rodas and Sara Magliacane. Identifiable Markov switching models with instantaneous effects and exponential families. ICML, 2026.
SWJ
Lorenzo Bertolini*, Roel Hulsman*, Sergio Consoli, Antonio Puertas Gallardo and Mario Ceresa. On general and biomedical text-to-graph large language models. Semantic Web Journal, 2025.
COPA'25
Roel Hulsman, Valentin Comte, Lorenzo Bertolini, Tobias Wiesenthal, Antonio Puertas Gallardo and Mario Ceresa. Conformal risk control for pulmonary nodule detection. COPA, 2025.
EU Technical Report
Valentin Comte, Lorenzo Bertolini, Roel Hulsman, Sergio Consoli, ... and Tobias Wiesenthal. AI-driven innovation in medical imaging. Technical Report, Publications Office of the European Union, 2025.
TEXT2KG @ESWC'24
Lorenzo Bertolini, Roel Hulsman, Sergio Consoli, Antonio Puertas Gallardo and Mario Ceresa. On constructing biomedical text-to-graph systems with large language models. TEXT2KG@ESWC, 2024.
Master Thesis
Roel Hulsman. Distribution-free finite-sample guarantees and split conformal prediction. Master's thesis, University of Oxford, 2022.