基于数据挖掘的运动负荷量监测方法  

Exercise load monitoring method based on data mining

作　　者：姜佳君[1] JIANG Jia-jun(Shaanxi University of Chinese Medicine,Xianyang 712046,Shaanxi Province,China)

机构地区：[1]陕西中医药大学,陕西咸阳712046

出　　处：《信息技术》2023年第2期91-96,101,共7页Information Technology

摘　　要：利用当前方法对运动负荷量进行监测时,未对运动员的心率信号降维处理,存在Kappa系数低、信号信噪比低和监测性能差的问题,为此,提出基于数据挖掘的运动负荷量监测方法。构建半监督式运动负荷邻域图,提取运动员的心率信号,完成数据挖掘;在半监督正则化方法的基础上,降维处理运动员的心率信号,保证数据计算的准确性;去噪低维信号,预处理提取的特征;通过支持向量机分类心率信号特征,实现运动负荷量监测。实验结果表明,所提方法的Kappa系数高、信号信噪比高、监测性能好。When using the current method to monitor the exercise load,there is no dimension reduction processing of athletes’heart rate signal,which has the problems of low Kappa coefficient,low signal-to-noise ratio and poor monitoring performance.Therefore,an exercise load monitoring method based on data mining is proposed.The semi-supervised exercise load neighborhood graph is constructed to extract the athlete’s heart rate signal and complete data mining.Based on the semi-supervised regularization method,the dimension of the athlete s heart rate signal is reduced to ensure the accuracy of data calculation,denoise the low-dimensional signal,and preprocess the extracted features.Finally,the support vector machine is used to classify the characteristics of heart rate signal to realize the monitoring of exercise load.The experiment results show that the proposed method has high Kappa coefficient,high signal-to-noise ratio and good monitoring performance.

关 键 词：数据挖掘 运动负荷量 SVM 去噪 特征提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]