Abstract
Aim/Background: Traditional classroom layout strongly affects student comfort, posture, attention, and learning. Thus, this study examines how machine learning combined with ergonomic principles can optimize classroom designs to better meet students’ physical and environmental needs.
Methods: Quantitative ergonomic and environmental data, such as anthropometrics, lighting, temperature, desk spacing, posture, and discomfort, were collected across all educational levels. Supervised models (Decision Trees, SVMs, Random Forests) predicted student discomfort, K-Means identified ergonomic zones, and a Genetic Algorithm optimized seating arrangements within ergonomic constraints.
Results: Among the evaluated models, the Random Forest classifier achieved the best predictive performance with an accuracy, precision, recall, F1 score and MSE values which are 85%, 0.86, 0.88, 0.87 and 0.13 respectively, outperforming Decision Trees and SVM. The optimized classroom layouts generated through GA-based optimization demonstrated measurable improvements in predicted comfort, posture alignment, and spatial efficiency compared to traditional layouts.
Conclusion: The integration of machine learning and ergonomics provides a robust, data-driven framework for designing adaptive and student-centered classroom layouts. The proposed approach bridges the gap between human-centered design and intelligent systems, offering scalable and practical solutions for improving physical learning environments in educational institutions.

Key words: Ergonomics, Machine Learning, Educational Environment, Spatial Optimization.