Solving university course timetabling problem using localized island model genetic algorithm with dual dynamic migration policy

Abstract

The University Course Timetabling Problem (UCTP) is a scheduling problem of assigning a teaching event in a certain time and room by considering the constraints of university stakeholders such as students, lecturers, and departments. This problem becomes complicated for universities with a large number of students and lecturers. Moreover, several universities are implementing student sectioning, which is a problem of assigning students to classes of a subject while respecting individual student requests, along with additional constraints. Such implementation also implies the complexity of constraints, which is larger accordingly. However, current and generic solvers have failed to meet the scalability and reliability requirements for student sectioning UCTP. In this paper, we introduce the localized island model genetic algorithm with dual dynamic migration policy (DM-LIMGA) to solve student sectioning UCTP. Our research shows that DM-LIMGA can produce a feasible timetable for the student sectioning problem and get better results than previous works and the current UCTP solver. Our proposed solution also consistently yield lower violation number than other algorithms, as evidenced by UCTP benchmark experiment results.

Genetic Algorithm for Exam Timetabling Problem - A Specific Case for Japanese University Final Presentation Timetabling

“…The system was designed to allocate students and supervisors in a more efficient way to reduce the number of rooms and time consumption. In 2020, Gozali, et al [15] attempted to solve the university course timetabling by using localized island GA with dual dynamic migration policy (DM-LIMGA). The results show that the proposed algorithm can produce a feasible timetable in student sectioning problem with a better result than previous works.…”

“…For example, there is a problem with 5 sessions, [1,2,3,4], [5,6,7], [8,9,10,11], [12,13,14] and [15,16], and there are two parent chromosomes, Assuming that session [8][9][10][11] and [15,16] are selected. The session [8][9][10][11] from parent chromosome 1 and session [15,16] from parent chromosome 2 will be swapped. Similarly, the session [15,16] from parent chromosome 1 and session [8][9][10][11] from parent chromosome 2 will be swapped as well.…”

“…The session [8][9][10][11] from parent chromosome 1 and session [15,16] from parent chromosome 2 will be swapped. Similarly, the session [15,16] from parent chromosome 1 and session [8][9][10][11] from parent chromosome 2 will be swapped as well. However, since the length of session [15,16] is shorter than that of session [8][9][10][11] and in parent chromosome 2, there are not enough blank timeslots beside the session [15,16].…”

“…Similarly, the session [15,16] from parent chromosome 1 and session [8][9][10][11] from parent chromosome 2 will be swapped as well. However, since the length of session [15,16] is shorter than that of session [8][9][10][11] and in parent chromosome 2, there are not enough blank timeslots beside the session [15,16]. Therefore, the session [12][13][14] in the parent chromosome 2 will be moved two timeslots forward to vacate enough position for session [8][9][10][11].…”

“…16,0,0,0,0,0,0,0,0, 9,8,10,11,0,0,0,0,0,0, 13,12,14,0,0,0,0,0,0,0, 5,6,7,0,1,2,3,4,0,0] and [0,0,0,0,0,0,0,0,0,0, 8,9,10,11,0,1,2,3,4,0, 0,0,00,0,0,0,0,0,0, 7,6,5,0,0,12,13,14,15,16] …”