PhD Positions in Robust Optimization, Machine Learning, and Statistics at Centrum Wiskunde & Informatica

Centrum Wiskunde & Informatica (CWI) in Amsterdam is offering two fully funded PhD positions in the Stochastics and Machine Learning research groups, supervised by Professor Bart Van Parys. These positions are part of the NWO Vidi project “Robust Decisions with Wild Data,” focusing on the development of new mathematical and algorithmic foundations for robust optimization, machine learning, and statistics. The research addresses the challenges of making reliable decisions with algorithms in the presence of messy, heavy-tailed, adversarial, or dynamically collected data. Key topics include robust optimization under uncertainty, anytime-valid and robust statistics, regret minimization, online learning, learning-augmented algorithms, and optimization for distributionally robust and nonconvex models. The project aims to advance methodologies that provide certified performance beyond worst-case scenarios, particularly in the context of wild data sets. Applicants should hold a Master’s or MPhil/Research Master’s degree in mathematics, statistics, operations research, or a closely related field, with an excellent academic record (top 10% of class). Strong English proficiency, academic writing and presentation skills, and experience with scientific programming (Python, Julia, Matlab) are valued. The positions are fully funded, with a gross monthly salary starting at €2,968 and rising to €3,801 over four years, plus holiday and year-end bonuses, flexible scheduling, and expat housing support. CWI is the Dutch national research institute for mathematics and computer science, located at Science Park Amsterdam. The institute is internationally oriented, with a diverse workforce and a commitment to inclusion. Applications should include a detailed CV, motivation letter, MSc course list and grades, Master’s thesis, and publications. The deadline for full consideration is April 1, 2026. For more information, contact Professor Bart Van Parys at [email protected] or visit the CWI website.