Time table scheduling using Genetic Algorithms employing guided mutation

Genetic Algorithms, a class of evolutionary optimization techniques offer benefits of being probabilistic, requiring no auxiliary knowledge in comparison to conventional search methods such as calculus based, enumerative and random strategies. This paper discusses a Genetic Algorithm based university time table scheduling algorithm satisfying constraints that avoid clash of faculty, class room slots, etc. The paper exploits the rank based selection scheme to ensure that the time table schedule generated is the feasible global optima as opposed to the stagnant solution setup associated with roulette selection scheme. An application specific encoding structure, rank based selection of time-table schedules and single point crossover to explore new and fitter schedules is used in the proposed algorithm. The proposed guided mutation operator helps in convergence as a result of the increased constraint satisfaction rates and hence better fitness values.