Optimisation of Students' Examination Seating Arrangements Using Simulated Annealing

Abstract

Exam seating arrangements must be carefully designed to reduce academic dishonesty. Many institutions make deliberate seat assignments to minimise the possibility of proximity-based collusion. However, typical methods frequently fail in large student populations, stressing the importance of optimisation strategies. This work investigates the use of Simulated Annealing (SA) to generate near-optimal seating arrangements, considering limitations such as students' courses, departments, and academic levels. The SA algorithm is constructed and assessed using synthetic data for a 40-student examination hall. The results indicate that SA effectively reduces adjacency violations and has practical applications in institutional exam seating arrangement planning.