Clustering of Pre-Service Teachers by Stress Levels Using Machine Learning Techniques

Teacher professional experience training is crucial in developing professional educators. Consequently, universities nationwide that offer education degrees include teacher professional experience training in their programs. This study aims to cluster of pre-service teachers by stress levels from teacher professional experience training. The dataset includes 208 samples from the Faculty of Education at Nakhon Ratchasima Rajabhat University, Thailand, during the 2022 academic year. This work presents a clustering of pre-service teachers using three machine learning algorithms: K-means Clustering, Hierarchical Clustering, and Spectral Clustering. A comparison of clustering performance revealed that Spectral Clustering achieved the highest Silhouette Coefficient at 0.3741. Two clusters were identified: one comprising 171 members with low-to-moderate stress levels and another with 37 members experiencing high stress levels. These findings suggest the need for targeted interventions and personalized support to address the varying stress levels among pre-service teachers. Future research should incorporate longitudinal studies to monitor changes in stress levels over time and evaluate the long-term impact of stress management interventions.