Combination of Graph Heuristic with Hill Climbing Search for Solving Capacitated Examination Timetabling Problem
Examination timetabling problem is a nontrivial task due to its NP-hard nature as well as the involvement of numerous constraints. Moreover, solving capacitated examination timetabling is more challenging compared to uncapacitated one. This paper implements graph heuristic with hill climbing search...
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/11072/ http://umpir.ump.edu.my/id/eprint/11072/ http://umpir.ump.edu.my/id/eprint/11072/1/Combination%20of%20Graph%20Heuristic%20with%20Hill%20Climbing%20Search%20for%20Solving%20Capacitated%20Examination%20Timetabling%20Problem.pdf http://umpir.ump.edu.my/id/eprint/11072/7/fskkp-ashis%20kumar%20mandal-combination%20of%20graph%20heuristic.pdf |
| Summary: | Examination timetabling problem is a nontrivial task due to its NP-hard nature as well as the involvement of numerous constraints. Moreover, solving capacitated
examination timetabling is more challenging compared to uncapacitated one. This paper implements graph heuristic with hill climbing search to solve the capacitated examination timetabling considering partial examination assignment concepts. The algorithm starts with ordering all the exams according to graph heuristic approach and then partial exams are considered for construction. Afterwards, the qualities of these exams are improved using hill climbing search. The entire process continues until scheduling all the exams. The effects of different graph heuristic orderings and exam assignment values on the quality of the solutions are studied. We test the proposed approach on ITC2007 benchmark exam datasets which contains highly constraint capacitated datasets. Experimental results reveal that the approach is able to produce quality solutions for all datasets and competitive results with competition results. |
|---|