Iterated Heuristic Algorithms for the Classroom Assignment Problem

We tackle the classroom assignment problem in a large University with the objective of minimising the total distance between all classrooms assigned to teaching activities in the same course. Additional requirements that should be satisfied include: making an efficient utilisation of the space, satisfying room preferences and complying with other administrative requirements. We present two iterated heuristic approaches, each one consisting of an iterative resolution of an assignment problem (the classical assignment problem in the first approach and the bottleneck assignment problem in the second approach) and a third algorithm based on the Variable Neighbourhood Search (VNS) meta-heuristic. We also present and discuss experimental results using real-world data from three consecutive academic sessions.