Physical Fitness Evaluation of College Students at the Stage of Physical Exercise Behavior Based on Bayesian and Data Mining

It is difficult for the traditional physical fitness evaluation methods to dig useful information from massive data, and the accuracy of physical fitness evaluation is low. Therefore, this study proposed a physical fitness evaluation method in the stage of physical exercise behavior based on Bayesian and data mining for the college students. The purpose was to set the association rules of exercise behavior stage, to mine the association mapping relationship in the data set by using frequent itemsets, to build a regression model and to select the best physical variables in the exercise behavior stage. The frequent itemset was used to eliminate the redundancy of physical fitness data, the causal relationship was used to sort the physical exercise behavior stage, and the transformation of physical fitness evaluation index system was realized through Bayesian network topology to realize the physical fitness evaluation in the physical exercise behavior stage. The experimental results showed that the accuracy of this method was as high as 97.62%, and the recall rate of fitness data evaluation was as high as 99.3%. At the same time, the effect of fitness evaluation was better in a short evaluation time.