Multiple Data Analysis of College Students' Physical Health Based on Big Data and Self-Organizing Map

Abstract

College Students' physical health data has the characteristics of high dimension, and the amount of data is particularly large. Big data method and self-organizing feature mapping network (SOM) method have unique advantages and visualization characteristics for processing high-dimensional mass data, so they become important tools for pattern recognition and visualization analysis of big data. Taking Zhejiang University of Commerce students' physical health data as an example, this paper analyzes the regional characteristics of influencing factors and explanatory factors of students' physical health with big data and visual SOM method. Results the analysis showed that body weight and BMI index had regional consistency, which were the most important factors affecting students' physical health, and were also the main explanatory variables of students' physical health status. The regional difference of physical health of girls is relatively large, while that of boys is small. The clustering characteristics of students' physical health indicators also have regional consistency. This paper demonstrates the rationality of big data method and self-organizing feature mapping network method in pattern recognition and visual analysis of physical health data. The results provide a certain reference value for the analysis of physical big data.