Nurse rostering, focus on models or algorithms?
Greet Vanden Berghe

Nurse rostering is a particularly challenging optimisation problem that includes many human factors. Nurse rostering has been subject of many research projects, which have resulted in good algorithms for producing rosters that satisfy many requirements. While automated rostering tools based on these algorithms are fast, they are not always perceived as superior to manual rostering practice. The talk will address a few specific characteristics of nurse rostering problems for which, for the time being, only scattered models exist.

Biography: Greet Vanden Berghe is Associate Professor Computer Science at KU Leuven, Faculty of Engineering Technology. She currently chairs the Technology Cluster Computer Science and leads CODeS research group in Gent. Her research focuses on combinatorial optimisation and automated decision support:

  • developing efficient heuristic approaches to real life problems, mainly in the health care and logistics domains,
  • developing effective adaptive heuristics, e.g. hyperheuristics, to a wide range of combinatorial optimisation problems,
  • developing generic approaches to the class of structured problems, i.e., problems incorporating both components from scheduling and routing, rostering and scheduling, etc.