24th Conference of the International Federation of Operational Research Societies
Abstract Submission

4038. Revisiting Local Search for Curriculum-Based Course Timetabling

Invited abstract in session TF-43: Metaheuristics for Scheduling and Timetabling, stream Metaheuristics.

Tuesday, 16:15-17:45
Room: JUR – Seminar-Raum 63

Authors (first author is the speaker)

1. Andrea Schaerf
Polytechnic Department of Engineering and Architecture, University of Udine
2. Sara Ceschia
Polytechnic Department of Engineering and Architecture, University of Udine
3. Francesca Da Ros
Department of Medical Oncology, Centro di Riferimento Oncologico IRCCS
4. Luca Di Gaspero
DPIA, Università degli Studi di Udine

Abstract

We address the Curriculum-Based Course Timetabling problem, a well-known benchmark in educational timetabling, and we propose a Simulated Annealing approach based on a new neighborhood, LectureKick, extending the classic LectureMove to allow more flexible lecture relocation. The method includes probabilistic biases targeting soft constraints and a cut-off mechanism in the cooling schedule. After parameter tuning on artificial datasets, experiments on real-world benchmarks show that the approach outperforms previous local search methods and competes with state-of-the-art algorithms.

Keywords

Status: accepted

Back to the list of papers