Authors : Ariff Md Ab Malik, Masri Ayob Abdul Razak Hamdan and Zulkifli Ahmad
Abstract: The method of focusing only on intensification or diversification within a Tabu Search algorithm may jeopardize the possibility in finding better quality of solutions. Thus, the decision to switch between intensification and diversification during solving process may enhance the overall performance of the search. This research proposes an adaptive switcher mechanism in Integrated Two-stage Multi-neighbourhood Tabu Search (ITMTS) to decide when to switch from intensification to diversification. The trigger value is calculated from the mean plus the standard deviation of the qualities of some recently accepted solutions. When the search consecutively generates bad (or same) quality solutions, it will automatically switch from intensification to diversification. The search will switch back to intensification phase after performing the shaking process. The technique is tested on examination timetabling datasets and demonstrates the effectiveness of adaptive switcher mechanism.
Ariff Md Ab Malik, Masri Ayob Abdul Razak Hamdan and Zulkifli Ahmad, 2013. An Adaptive Switcher Mechanism of Integrated Two–Stage Multi–Neighbourhood Tabu Search for University Examination Timetabling Problems. Research Journal of Applied Sciences, 8: 152-160.