To search, Click below search items.

 

All Published Papers Search Service

Title

A Comparison between Hybrid Population-based Approaches for solving Post-Enrolment Course Timetabling Problems

Author

Ghaith M. Jaradat, Masri Ayob

Citation

Vol. 11  No. 11  pp. 116-123

Abstract

In this study, we apply our Elitist-Ant System, Big Bang-Big Crunch and Scatter Search heuristics to solve two post-enrolment course timetabling problems (first and second international timetabling competitions) and to compare their performance and consistency. The approaches mainly focus on employing the elite pool and solution combination strategies. Both strategies provide deterministic search guidance by maintaining a balance between diversity and quality of the population. This is achieved by a dynamic changing of the population size and, the utilization of elite solutions and a probabilistic selection procedure in generating good quality and diversity solutions. Experimental results showed that our hybrid approaches produce good quality solutions, and outperforms some best known results reported in the literature including population-based algorithms. In term of solutions’ quality, the Scatter Search ranked first and followed closely by the Elitist-Ant System and Big Bang-Big Crunch heuristics.

Keywords

Elitist-Ant System, Big Bang-Big Crunch, Scatter Search, post-enrolment course timetabling problem, elite pool, solution combination

URL

http://paper.ijcsns.org/07_book/201111/20111119.pdf