Exploiting Machine Learning Techniques to Discover Physical Fitness Patterns Among Diverse Clusters of Female College Students

To target improving the physical fitness for female college students and to meet the individualized needs of different subgroups. A total of 6987 female college students from a teacher training college were assessed for physical fitness. Their fitness indices were analyzed using K-means clustering and GBDT regression algorithms. Based on physical fitness profiles, students were grouped into four categories: Category I (Normal body shape, average fitness and athletic ability); Category II (Normal body shape, good fitness and athletic ability); Category III (Short, poor fitness and athletic ability); Category IV (Overall excellent fitness). Physical development and factors influencing the 800-meter run vary across groups of female college students. This study provides insights for more targeted interventions to improve specific physical fitness.