Improvement of Key Feature Mining Algorithm for Sports Injury Data Based on LOF Enhanced K-Means and Sparse PCA

Abstract

Sports injury not only affects the health of athletes, but also has a negative impact on their sports performance and competitive level. By mining the key features of sports injury data, we can identify the key factors that affect athletes' performance, so as to improve sports performance and competitive level. Therefore, this paper proposes an improved key feature mining algorithm for sports injury data based on LOF enhanced k-means and sparse principal component analysis. The basic probability assignment method of attribute weight is used to assign the damage data, which provides a neat and consistent data basis for the subsequent key feature mining of sports injury. The K-means algorithm improved by LOF algorithm is used to classify the assignment results and divide the sports injury data. PCA is used to reduce data dimensions, simplify redundancy, and enhance the independence of sports injury data features. Using reweighted sparse PCA to realize key feature mining of sports injury data. The experimental results show that the proposed method can accurately capture the essential differences between non sports injury data and sports injury data, and accurately divide non sports injury data and sports injury data. At the same time, the average absolute percentage error and root mean square error of the assessment accuracy of injury factor assignment are both lower than 0.1, and the DBI values of all samples are not more than 0.13, it can effectively mine the key features of sports injury data.

翻译

搜索

复制