A time-predefined local search approach to exam timetabling problems

Abstract

In recent years the processing speed of computers has increased dramatically. This in turn has allowed search algorithms to execute more iterations in a given amount of real-time. Does this necessarily always lead to an improvement in the quality of final solutions? This paper is devoted to the investigation of that question. We present two variants of local search where the search time can be set as an input parameter. These two approaches are: a time-predefined variant of simulated annealing and an adaptation of the "great deluge" method. We present a comprehensive series of experiments which show that these approaches significantly outperform the previous best results (in terms of solution quality) on a range of benchmark exam timetabling problems. Of course, there is a price to pay for such better results: increased execution time. We discuss the impact of this trade-off between quality and execution time. In particular we discuss issues involving the proper estimation of the algorithm's execution time and the assessment of its importance.

A graph-based hyper-heuristic for educational timetabling problems

“…The state-of-theart approaches in educational timetabling usually employ specially tailored heuristic/metaheuristic approaches (e.g. Carter, Laporte and Lee, 1996, Di Gaspero and Schaerf, 2000, Caramia, Dell'Olmo, and Italiano, 2001, Burke and Newall, 2002, Merlot et al, 2002, Socha, Knowles and Sampels, 2002, Asmuni, Burke and Garibaldi, 2004, Abdullah et al, 2004, Burke, Bykov et al, 2004. These approaches invest a significant amount of development effort in the production of "fine tuned" techniques that are specially built for the particular problems in hand.…”

Simulated Annealing with Deterministic Decisions

“…Another variation is the old bachelor [3] algorithm, which is similar to the threshold acceptance algorithm, except that if a solution is not accepted the threshold is increased. The degraded ceiling algorithm [4] is yet another variant with absolute bound threshold that decreases by time.…”

Investigating a Hybrid Metaheuristic for Job Shop Rescheduling

“…The simulated annealing and great deluge algorithms are then applied respectively in the second phase by using the best antibodies selected in the first phase (final population) as initial solutions. In this research, we are using the parameters adapted from [3]. The aim is to investigate if we can improve the fitness of the antibodies developed in the final population in Phase I as both local search methods have been known to produce good results for other scheduling problems such as examination timetabling (e.g.…”

“…The aim is to investigate if we can improve the fitness of the antibodies developed in the final population in Phase I as both local search methods have been known to produce good results for other scheduling problems such as examination timetabling (e.g. [3,8]). …”

“…This estimated quality represents the final estimated fitness value of an antibody. The boundary to the fitness of each antibody known as boundary is decreased by a decreasing rate, β [3] which is defined as follows:…”