Tabu exponential Monte-Carlo with counter heuristic for examination timetabling

In this work, we introduce a new heuristic TEMCQ (Tabu Exponential Monte-Carlo with Counter) for solving exam timetabling problems. This work, an extension of the EMCQ (Exponential Monte-Carlo with Counter) heuristic that was originally introduced by Ayob and Kendall. EMCQ accepts an improved solution but intelligently accepts worse solutions depending on the solution quality, search time and the number of consecutive non-improving iterations. In order to enhance the EMCQ heuristic, we hybridise it with tabu search, in which the accepted moves are kept in a tabu list for a certain number of iterations in order to avoid cyclic moves. In this work, we test TEMCQ on the un-capacitated Carter's benchmark examination timetable dataset and evaluate the heuristic performance using standard proximity cost. We compare our results against other methodologies that have been reported in the literature over recent years. Results demonstrate that TEMCQ produces good results and outperforms other approaches on several benchmark instances.