Cross-Disciplinary Applications of Evolutionary Computation in STEM Education Analytics

Abstract

Evolutionary Computation (EC) methods such as Genetic Algorithms and Particle Swarm Optimization are gaining traction in educational research, particularly across STEM disciplines. This study investigates EC applications in curriculum design, predictive modeling, and student clustering within science, mathematics, and engineering education analytics. Using data from 3,000 students across five institutions, our EC-enhanced models outperform baseline logistic regression and k-means in accuracy (+8%) and cluster purity (+12%), and reduce curriculum testing cycles by 25%. We discuss implications for scalable, adaptive learning systems informed by interdisciplinary analytics.