Web based personalized university timetable for UiTM students using genetic algorithm / Mohd Radhi Fauzan Jamli and Ahmad Firdaus Ahmad Fadzil

The research endeavors to develop a web-based system specifically designed to create personalized university timetables for Universiti Teknologi MARA (UiTM) students using genetic algorithms, aiming to address the urgent need for a customizable timetable solution catering to the diverse scheduling requirements of both repeater and non-repeater students while optimizing course group selection to minimize conflicts and enhance scheduling flexibility. The complexity of timetable generation stems from the varied course groupings and scheduling constraints inherent in UiTM's curriculum, leading to challenges for students, particularly repeaters, in enrolling in courses across different semesters and groupings, resulting in conflicts and inefficiencies. Traditional methods of timetable generation lack the adaptability needed to tackle these complexities, necessitating the development of an innovative solution. The proposed approach utilizes genetic algorithms to dynamically produce optimized timetables based on individual student needs, with real-time data scraping from 'iCRESS' ensuring the system stays up to date with the latest course information for accurate timetable generation. Within the genetic algorithm framework, each timetable is represented as a chromosome, forming a population of potential timetables refined through successive generations by genetic operators like crossover and mutation. Student input initiates the process, with user interaction allowing for timetable customization based on personal preferences. Extensive experimentation with genetic algorithm parameters has yielded promising results, notably a parameter set (population size = 12, generation size = 30, mutation rate = 0.2) demonstrating robust performance, achieving optimal timetables with swift convergence and minimal conflicts. This configuration excelled in efficiency and scalability, offering a viable solution for timetable generation at scale. Future work entails enhancing system robustness through comprehensive contingency planning, real-time data integration, and algorithmic optimization, with a focus on refining the genetic algorithm and exploring parallel processing techniques to further enhance efficiency and scalability.