ANALYSIS OF COLLEGE STUDENTS' PHYSICAL FITNESS TEST DATA AND RESEARCH ON PHYSICAL EDUCATION CURRICULUM REFORM

Abstract

This study collected standardized physical fitness test data from 28,290 college students and conducted an in-depth analysis using a complete data preprocessing pipeline and machine learning algorithms. First, missing values were imputed and outliers were processed, followed by standardization. Principal component analysis (PCA) was then applied for dimensionality reduction, and the K-means clustering algorithm was used, with the optimal number of clusters (K=2) determined by the silhouette coefficient method. The data analysis revealed significant group differences in students' physical fitness, particularly in strength and cardiorespiratory endurance. Based on these findings, a 16-week teaching reform experiment was conducted at Guangzhou Huashang College, incorporating weak test items into the curriculum. The results showed that the pass rate for male students' pull-ups increased by 23.34 percentage points (p<0.01), and the standing long jump improved by 25 percentage points (p<0.01). Female students also demonstrated significant improvements in related indicators. This study validates the effectiveness of data-driven physical education reform and provides actionable pathways for college physical education programs.