Reducing Class-Scheduling Conflicts Using Linear Programming

摘要

The scheduling of college courses within a given semester plays a crucial role in a student's daily life, educational and professional pursuits, and future successes. Inadequate class scheduling has the potential to delay student graduation. There is no standardized system for scheduling classes for the Department of Civil and Environmental Engineering at University of Hawaii (UH) at Manoa, which generates numerous class conflicts every semester. This article proposes an optimization model combining visual modeling and linear programming software that can be used to identify and eliminate conflicts when scheduling both undergraduate-level and graduate-level courses within an individual department. Constraints for this model were created to ensure that prerequisites are enforced, overlapping classes are minimized or eliminated, and maximum numbers of classes are available to the students every semester. Schedules for eight semesters were tested by the model, and then compared to the newly optimized version to determine if class conflicts could be reduced. Results showed that the proposed modeling methodology improved every semester's schedule by an average of 83.46%, with a minimum confidence of 95.14%, thereby attesting to the validity of this model, allowing students more class options and the ability to graduate sooner. (c) 2017 American Society of Civil Engineers.