The Study And Implementation Of Decision Support System Based On Data Mining

China's university students in physical training mainly base on classroom study, and take exercise in off-hour to be subsidiary, basically in a state of disorder, there are many unreasonable factors, unable to meet the needs of college students to improve physical fitness. With the computer science and technology development, especially in the development of decision support systems, decision support system in the field of sports training has been applied, with the development of data warehouse and data mining technology ,they were applied to them.This paper introduces the sports training aids decision support system, which uses data warehouse and data mining technology。The college students' different aspects are in organic integration, and advanced training methods are used in sports management, combined with users' input to have a reasonable plan of sports training. The system is based on the Nanchang University students in school sports and physical examination results, and so on to build a data warehouse cube, and data mining techniques are applied to them.Mining association Rules algorithm and decision tree classification algorithm are given a detailed analysis and research.First, its classic apriori algorithm is made a appropriate improvement, enhancing the speed of reading the effective data in the database, simultaneously reducing scanning some non-essential data, it can produce the frequent item set more quickly. The two aspects can apparently improve the efficiency of the data mining, especially for the magnanimous data. Secondly, the ID3 decision tree algorithm is improved,and a model is built based on the method of mining model, a reduction of contributions on the selection of properties when calculating the gain value, and an important attribute is set, which solves a problem that root node tend to be the more property value, and set a threshold of pruning to limit the growth of the tree entirely, to increase the speed at the same time, to have a reasonable structure of the decision tree.Mining association rules model was established based on this algorithm which was improved, meanwhile using sqlserver 2005 Analysis Service to establish a mining model based on multi-dimension data set, the rules which were produced by the two models are used to assist the sports teaching, and the strong association rules were imported into the knowledge base used as accord to produce athletic training plan.