Thesis (BS Applied Mathematics) -- University of the Philippines Mindanao, 2009

Nurse scheduling problem is the allocation of nursing staffs to different shifts while satisfying the different constraints. Genetic algorithm (GA) is a population based optimization method that has already shown its superior performances in scheduling problems. This study formulated a GA-based method for the nurse scheduling problem of Davao Medical Center. First, nurse aid and nurse schedules were represented as binary matrices. Then a minimization fitness function penalized a schedule for every violated constraint. Parameters were also set to 10000 maximum iterations, 20 individuals in the population, 100% occurrence rates for crossover and mutation processes and the mutation rate of 2%. The initial population includes two feasible individuals while the other 18 individuals were results of random plotting. Elitism method in selecting individuals that could go on to the next generation was done. Also, algorithm for uniform crossover and swap mutation processes were made. The formulated GA produced good nurse aid and nurse schedules. These schedules were further compared to the schedules of the existing manual and goal programming (GP) based methods. As compared to the manually made schedules, the best fit nurse aid and nurse schedules produced by GA were more optimal. In nurse aid scheduling, GP performed better than GA. However, in nurse scheduling, GA performed better than GP. Therefore, GA created schedules, which were as good as the schedules of the existing methods