Ann Nowé

The Artificial Intelligence Lab in the Department of Computer Science at Vrije Universiteit Brussel is offering a PhD position within Confident Data-Driven Decision Support (CoRDS) , a new European Doctoral Network funded under the Marie Skłodowska-Curie Actions. The project sits at the intersection of machine learning and operations research, with a strong focus on learning methods for optimization problems that must work across similar problem instances. The research topic is explainable, across-instance learning . You will investigate how reinforcement learning, combined with structured neural networks such as graph neural networks, can improve scheduling and other combinatorial optimization problems. A central challenge is how to transfer knowledge between related instances, make learned policies robust to concept drift, and understand how these methods generalize across optimization settings. Hosted at VUB and supervised by Prof. Ann Nowé , the position is part of a multidisciplinary, international network of doctoral candidates and supervisors. You will collaborate with the University of Vienna as an academic partner and with OMP as an industry partner, gaining exposure to both academic and applied research environments. The role includes developing reinforcement learning methods for structured optimization, creating explainability techniques that make solution strategies understandable to practitioners, and contributing to a public benchmark based on an optimization problem from OMP. The project also involves research dissemination through leading conferences and journals, as well as participation in the network’s training and networking activities. Overall, this is a research-intensive PhD in AI for optimization, with international collaboration and planned secondments in both academia and industry.