High School Timetabling Using Local Search Genetic Algorithm (LSGA) Sondari Setia Rahayu, Yaya Wihardi, Rani Megasari, Wawan Setiawan
Department of Computer Science Education, Universitas Pendidikan Indonesia, Jl. Dr. Setiabudhi, No. 229, Bandung, West Java, Indonesia
*sondari[at]upi.edu
Abstract
School timetabling is an activity to arrange teachers schedules to teach in the assigned classes. This activity is being conducted every semester and takes a long time to be completed, especially in high school, resulting in lower work effectiveness. This issue is caused by difficulty in distributing the schedules according to the given constraints. Several studies have found that artificial intelligence can help to solve the timetabling problem. One technique that can be used is the Local Search Genetic Algorithm (LSGA). In this research, the implementation of LSGA was carried out to address timetabling problems in high school. The test was run over five parameters: population, number of generations, probability of mutation, computation time, and convergence. The test results indicated that the final fitness value in LSGA is 1 while the value of GA is 0.064. These results show that LSGA can afford a convergent solution to the timetabling problem in high school and more optimal than GA.
Keywords: Genetic algorithm- High School Timetabling Problems (HSTP)- Local Search Genetic Algorithm (LSGA)
