Closed-loop management mechanism and system prototype implementation of college students’ physical health empowered by artificial intelligence

This study proposes a closed-loop physical health management mechanism powered by artificial intelligence (AI) to address the issue that university students’ physical health levels are continuing to deteriorate while the conventional management approach stresses monitoring and ignores intervention. The school’s overall evaluation model is built by optimizing the Classification and Regression Tree (CART) decision tree method, and the algorithms for class-wide scenario diagnosis and student-specific situation diagnosis are created. The physical health groups’ features are mined using K-means + + clustering, and a five-in-one closed-loop system of “monitoring-evaluation-feedback-intervention-guarantee” is built. Based on the physical health data of a total of 1100 college students from three universities, this study divides the dataset into a training set, a validation set and a test set at a ratio of 7:1:2, and adopted the fivefold cross-validation method for model training. The system has undergone a 6 month practical application verification in three pilot universities, and an independent sample t-test was conducted with the control group. According to the data, the CART algorithm’s accuracy is 87.5%, which is noticeably higher than that of other algorithms. Following the implementation of the system, there is a 13.4% increase in the physical health test compliance rate and a 26.5% increase in the percentage of students who actively participated in sports. The closed-loop system offers a precise and scientific approach for managing physical health in universities and universities, effectively transforming physical health work from passive response to active prevention.