Managing faculty leave and lecture adjustments in higher education institutions is often a challenging administrative task, typically performed manually through ad hoc scheduling. Inefficient allocation of substitute lecturers may result in increased workload imbalance, decreased teaching quality, and disruption of the academic timetable. This paper proposes a predictive modeling approach, powered by data science techniques, to optimize lecture adjustments when faculty members are on leave. By analyzing historical leave records, teaching schedules, workload distributions, and subject expertise, the proposed system can forecast potential leave patterns and recommend the most suitable substitute lecturers. Machine learning algorithms such as decision trees, random forests, and gradient boosting are employed to build predictive models for identifying leave trends and substitution requirements. The outcomes demonstrate that a data-driven approach not only improves scheduling efficiency but also enhances fairness in workload distribution among faculty members. This study highlights the potential of predictive analytics in transforming traditional faculty leave management into an intelligent, automated, and scalable decision-support system for academic institutions.
Faculty leave management, lecture adjustment, predictive modeling, data science, machine learning, workload balancing, academic scheduling, decision support system.