Application research on long jump training and guidance based on data mining

Tao Lu¹, Xiaojun Zhang²

COMPUTER MODELLING & NEW TECHNOLOGIES 2013 17(5C) 173-178

¹Qingdao Vocational and Technichal College of Hotel Management, Qingdao, Shandong, postcode 266100
²Henan Institute of Education, Henan Zhengzhou 450000, China

This paper adopts the data ware house technology and data mining technology to establish a sports training assistant decision support system for long jump athletes. To make organic integration of the training factors for athletes, it applies scientific training theory and advanced training methods to the sports training management. We focus on the improvement of two classic data mining algorithms: association rules of Apriori and decision tree classification ID3. For Apriori algorithm, we improve the connections and pruning strategy when creating (k+1)-order frequent item set by k-order frequent item set, and the process pattern of transactions. For the defects of ID3 algorithm, we propose to reduce the computation of attribute gain selection when establishing the tree, and provide corresponding scheme to set the attribute importance. Then the actual examples are used to apply the improved models to the sports training assistant decision support system. The results show our algorithm improve the mining efficiency actually. The generated strong association rules which have higher association can be imported into knowledge library as important base for the sports training schemes.