In this paper a new student sectioning method based on fuzzy clustering is presented. One of the less studied sub-problems of timetabling is student sectioning. This problem is due to courses, which involve a large number of students. We concentrated on initial student sectioning to create a section conflict graph prior to timetabling. Our aim is to allocate students to course sections so as to satisfy the following criteria: 1. Student course selections must be respected. 2. Section enrollments should be balanced, i.e. all sections of the same course should have roughly the same number of students; 3. Student schedules in each section would be the same as each other, as much as possible. In the proposed method, at first, with a Fuzzy C-means algorithm, students in large classes have been classified; then this clustering is evaluated with a fuzzy function, according to some criteria: clusters centers distance, clusters density, and size of clusters. Each student has a feature vector in fuzzy classifier. Taken curses of each student are its features. Based on the proposed fuzzy function and with an exhaustive search, the best features are selected. The best classification of students is corresponds with these selected features. The simulation results show that this method has fewer conflicts with respect to an entrance year based sectioning.

Keywords