
Individual factor analysis of wrestler’s performance based on SVM
Author(s) -
Naidan Xu,
Zhao Ling,
Zhengzhi Wu
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1941/1/012083
Subject(s) - support vector machine , principal component analysis , hyperplane , randomness , sample space , statistics , pattern recognition (psychology) , value (mathematics) , mathematics , artificial intelligence , computer science , sample (material) , decision boundary , chemistry , geometry , chromatography
Support vector machine (SVM) is a binary classification model, its algorithm is to map the sample data into a high-dimensional space. The hyperplane found in the high-dimensional space can accurately separate two kinds of data samples and maximize the interval. According to the problem of randomness and data imbalance of wrestlers' performance data, this paper proposes a new SVM model to analyze the personal factors of wrestlers' performance. Principal component analysis and chaos analysis are applied to optimize SVM model. Compared with BPNN, GM, ARIMA, LSSVM model, the contrast experiment result shows that while the number of samples increased to 300, the SVM model of the data processing time is less than 20 seconds. Operation time reduced minimum by 11% (compared with BPNN) and maximum by 60% (compared with GM). At the same time, the accuracy of sample analysis of SVM is about 90%-95%, which has a higher accuracy than other methods. Finally, the G-mean and F-value of the SVM model in the analysis of each principal component are significantly higher than the other four models. The G-mean value increases continuously when the imbalance is between 0.1 and 0.2, while the G-mean decreases gradually when the imbalance is greater than 0.324, indicating that the closer the decision boundary is, the closer the predicted results are to the true value.