Data Mining Applied In The Sunlight Sports System

In order to respond to national call to implement the Sunlight Sports practically,what to attract college students to go to the playground and nature under the sun for physical exercise, the state put forward the slogan of " Health, Sport, Sunlight, Future".At the same time the number of students increased rapidly by the scale of undergraduate enrollment continues to expand,the data what include attendance information and health data of students have increased greatly.So it is the inevitable requirement to manage the data well for the maturation at present.It has realized the autonomous and intelligence of Sunlight Sports through the introduction of punching the card by the use of high-tech means.It can guide and supervise students from the habit of physical exercise and obtain a good effect for Sunlight Sports. The system can analyze or manage the health data by the introduced technology and the system of Sunlight Sports.First of all, researched the classification algorithm of C4.5 decision tree,to improve the algorithm by the following aspects:(1) On the basis of the boundary method,improving the process of discretization, calculating the average of entropy and putting forward the way that use the average value as the optimal threshold, through that to improve the execution efficiency of discretization process.(2)To join the correlation coefficient as a parameter in the process to calculate the information gain rate by C4.5 algorithm, both considering the multivalued bias problem of information gain, and considering the influence of other attributes, making the selection of the test attribute more reasonable.(3)To simplify the improved calculation formula of information gain rate by maclaurin formula and transform the logarithmic operation to the operation of add, subtract and divide which is easy to computer would improve the algorithm performance further.Second, I designed and implemented the system of Sunlight Sports by analyze and research the actual situation of Sunlight Sports. The system of Sunlight Sports using SSH framework and includes five functional modules, respectively for personal center module, system management module,Knowledge management module, sports attendance module and help & feedback module.Finally determined the object, target,data collection and pretreatment of data mining, build the decision tree and generated the rule knowledge by the improved C4.5 algorithm, based on the real test data of physical health. Through the analysis to the decision tree, we can learn the important factors that affect physical health outcomes of students and the relationship between the indicators, so as to guide decision-makers adjust management means timely, in order to improve students’ body quality.