A fair course timetabling using genetic algorithm with guided search technique

Abstract

This paper presents a genetic algorithm (GA) combined with a guided search technique to solve the course timetabling problem. In the proposed method, the least selected and unused resources are keep tracked and stored in a data structures. As a guided or directed strategy, these data structures are then is used to improve the previously generated individuals by the genetic operators. The proposed algorithm also integrated a guided repair strategy and four neighborhood structures to make further improvements to the solution. Moreover, a penalty function that uses a sum of the square of the soft constraints violation was used to generate a fair and balance cost distribution in the timetable. Testing was performed using real-life dataset, and results indicate that the proposed genetic algorithm produces faster and better solutions compared to the classical genetic algorithm.