Timetabling the classes of an entire university with an evolutionary algorithm

Abstract

This paper describes extensions to an evolutionary algorithm that timetables classes for an entire University. A new method of dealing with multi-objectives is described along with a user interface designed for it. New results are given concerning repair of poor recombination choices during local search. New methods are described and evaluated that allow timetables to be produced which have minimal changes compared to a full or partial reference timetable. The paper concludes with a discussion of scale-up issues, and gives some initial results that are very encouraging.