Research On The Evaluation Method Of Sports Effects Based On Feature Selection

With the rapid development of Internet technology,data from various industries are accumulating explosively.The explosion in the fields of education,medical care,science and finance has promoted the development of data mining and other technologies in the era of big data.For example,in the financial field,big data and technologies have been fully penetrated into financial business scenarios,including credit,credit consumption rating,stock forecasting and information verification.Therefore,it can be seen that big data plays an important role in national development,and the research of big data technologies are the demands of the current era and the focus of attention from all walks of life.Along with the rapid development of data mining and machine learning,sports large data to development brings the challenge,the existing sports mainly focus on the data mining method to extract and construct the effective foundation of sports data characteristics,combined with the basic characteristics of the use of statistical methods to analysis and research of sports data,or use the early traditional approach to data mining.The mining of sports data cannot be carried out simply by using data statistics method.How to combine machine learning technology to effectively mine and analyze sports data,so as to provide beneficial suggestions for sports exercise,which is an urgent problem to be studied.Research on sports effect by feature selection algorithm is an efficient research on sports data mining.Evaluation of sports effect is analyzing and evaluating the influence of sports on physical indicators.A large number of literature studies have shown that most of the data sets used in the study of sports effect evaluation techniques are taken from the bulletin of national physique monitoring,and a small number of studies can be obtained by using questionnaire surveys and other means.Since national physical fitness monitoring is to monitor and analyze the overall condition of national physical fitness through sampling survey at the same time,the data finally obtained by this work is cross-sectional data,and the problem of using this data to conduct physical fitness research is often that the overall analysis of individual differences is not very good.In addition,national physical fitness monitoring takes a long time,and the data acquired in a short period of time are reused for research,resulting in low innovation in scientific research.Therefore,the acquisition and use of data sets is a problem that needs to be solved in the study of physical exercise effects.Focusing on the above-mentioned difficulties in the study of sports effects,in view of the limitations of existing data sets and traditional research methods,this article starts with data mining algorithms and constructs a sports effect evaluation database.Based on the idea of feature selection,elastic network algorithms,Random forest algorithm,carried out the research on the effect of sports on physical indicators.The main content of this paper includes:1.Build a sports effect evaluation database.Different from the Physical Fitness Research Institute based on the content of the national test report,this paper selected 785 teenagers as the research object to carry out the physical exercise research,collected the real training data,preprocessed,sorted and marked the data,and constructed the Sports Effect Datasets(SED).Compared with the existing data,the database proposed in this paper is more complex than the national physical fitness monitoring report,which effectively takes into account the differences of individual data.At the same time,the release of this database can provide data research basis for sports effect evaluation and even sports data mining in the future.2.Propose a sports effect evaluation algorithm based on elastic network.Compared with the traditional physique research method,the algorithm introduced machine learning algorithm and feature selection algorithm to guide the evaluation research of sports effect.To study sports effect evaluation,we add the elastic network algorithm for regularization optimization,which makes the study of physical fitness more scientific and can reveal the effects of sports as much as possible.The experimental results show that the selected features and Ground-truth utilization evaluation index analysis and research,compared with the baseline method,the algorithm has better accuracy.3.Propose a sports effect evaluation algorithm based on random forest.Compared with traditional research methods,this algorithm applies feature selection algorithm to study the influence of sports on fitness effectiveness.We use the information gain index to study the feature selection ability of the algorithm,which can scientifically and accurately obtain the influence degree of sports on different physical indicators,and comprehensively study the effect of sports.In this paper,we choose the training evaluation model of SED.Experimental results show that the proposed evaluation method has better performance and higher accuracy than the existing evaluation methods based on classical feature selection.