摘要
提出了一种基于最小二乘支持向量机(LS-SVM)的刘翔专项成绩预测的新方法。收集2000年至2007年59次刘翔110 m栏成绩组成整个数据集,前5次的成绩用来预测第6次的成绩,由前54次成绩建立LS-SVM成绩预测模型,运用建立的LS-SVM成绩模型预测最后5次的成绩。结果表明:提出的LS-SVM成绩预测方法是有效的。刘翔110 m栏成绩预测可为其训练乃至世界优秀跨栏运动员的训练提供理论参考。
The prediction of Liu Xiang＇s achievements can provide the theoretical reference for the training of Liu Xiang, as well as the other hurdle athletes. A new prediction approach for Liu Xiang＇s achievements is put forward by means of Least Squares Support Vector Machine （LS-SVM）. 59 achievements of Liu Xiang from year 2000 to 2007 were collected and construct the total data set. Achievements in the first 5 times were used to predict the achievement in the 6th times. The first 54 achievements were applied to build LS-SVM model for special achievements prediction, then, the prediction is conducted using achievements of last five times. The results showed that LS-SVM prediction method was a promising tool.
出处
《天津体育学院学报》
CAS
CSSCI
北大核心
2009年第4期330-333,共4页
Journal of Tianjin University of Sport
关键词
最小二乘支持向量机
刘翔
预测
成绩
Least Squares Support Vector Machine （LS-SVM）
Liu Xiang
prediction
achievements
