PhD position: Explainable, across-instance learning (MSCA Doctoral Network)

Doctorat / PhD Profil chercheur : First Stage Researcher (R1) Type de contrat : Temporary Date limite de candidature : 30 Nov 2026 - 18:00 (Europe/Brussels) Date de début : 1 Jan 2027 Programme UE : Horizon Europe - MSCA Département : Computer Science Domaine : Computer science » Informatics

The Artificial Intelligence Lab in the Department of Computer Science at Vrije Universiteit Brussel (VUB) is looking for a motivated PhD researcher in Computer Science. You will join Confident Data-Driven Decision Support (CoRDS) , a new Doctoral Network funded by the European Union under the Marie Skłodowska-Curie Actions (MSCA) Doctoral Networks programme (Grant Agreement No. 101227512). The position is hosted at VUB under the supervision of Prof. Ann Nowé, with the University of Vienna as academic partner and OMP as industry partner. Project How can reinforcement learning combined with e.g. graph neural networks solve scheduling problems faster and better? How can we transfer knowledge between similar problem instances? How can we better understand how reinforcement learning methods generalize across similar optimization problems? If these questions excite you, and you have a solid ML and/or OR background, we invite you to apply for this PhD topic on explainable, across-instance learning. You will work at the intersection of machine learning and operations research. You will develop learning-based methods that generalize across problem instances and can be explained to the people who use them. You will be part of a multidisciplinary, international team of doctoral candidates and supervisors across the network. Responsibilities You will: develop reinforcement learning methods based on structured neural networks for scheduling and other combinatorial optimization problems; design methods that transfer knowledge across problem instances and stay robust under concept drift; study and explain how reinforcement learning methods generalize in optimization settings; develop explainability techniques that make learned solution strategies understandable to practitioners; carry out a secondment at the University of Vienna with strong expertise in OR; carry out a secondment at OMP to apply and refine your research in an industry environment, and help develop a public benchmark based on an optimization problem from OMP; take part in the training and networking activities of the MSCA Doctoral Network; publish results in leading conferences and journals and present your work internationally. Profile We are looking for a candidate with: a Master’s degree in Computer Science or a related field (e.g. Mathematics, Engineering, Operations Research), preferably with a specialization in machine learning or optimization; a strong interest in reinforcement learning, structured neural networks and combinatorial optimization; good programming and computational skills; enthusiasm for research that links AI method development with industrial applications; a strong command of…