Statistical Analysis and Data Mining Combined Yoga Grade Prediction

Since 2002 all kinds of schools in China began to execute "National Student Physical Health Standard", in order to improve children and youth fitness and health. Huge amount of test data was accumulated year by year. However, students and teachers are losing their patience towards such a kind of test because test data are not utilized very well to improve their fitness and health. One important reason is that statistical analysis is not suitable for huge data set. On the contrary, data mining is considered a prospective method. In this paper, we try to make grade prediction based on physical health data to help student choosing a suitable PE class, combining statistical analysis and data mining methods. A process including correlation analysis, clustering and naive bayes based classifying is designed to fulfill this task. Experimental result shows that predicting accuracy is fairly good.