Research On Poverty Student Identification And Academic Early Warning Model Based On Data Mining

With the exponential growth of educational data,how to use these data to improve the quality of management decision-making has become one of the most challenging tasks in the current teaching work.In previous studies,the data of educational data mining mainly comes from the data of educational process or teaching content,and pays little attention to educational objects or educational organizers.Therefore,this paper chooses students as the research object,and carries on the identification and academic early warning of poor students through the data of family economic situation,daily consumption,learning situation,ideological state and so on.The main work of this paper is as follows:First of all,this paper makes an in-depth analysis of the current research situation at home and abroad from three aspects: data mining algorithm,identification of poor students and academic early warning,and points out some problems existing in the application of educational data mining.at the same time,in view of the shortcomings such as the singleness of data source and the unsatisfactory effect of the algorithm,this paper puts forward the model research on the identification and academic early warning of poor students,and briefly introduces the concepts and software involved in the research process.Secondly,according to the information sources such as national policy,social survey and expert analysis,this paper comprehensively considers the factors that affect the identification and study of poor students,determines the standards that affect the identification and study of poor students,and obtains data from the school data center.then the data are preprocessed with the techniques of data cleaning,data discretization and conceptual stratification,and the data are analyzed visually.Then,based on the analysis of the current mainstream algorithms,a classification algorithm for poor students’ identification and academic early warning is proposed.The information gain rate is selected as the classification standard to generate the decision tree,the pessimistic pruning algorithm is used to prune the decision tree,and the continuous correction of binomial distribution is introduced to reduce the error.and add the integrated algorithm to divide the data set into multiple balanced subsets for training respectively,and vote to select the model with high accuracy.Finally,the classification accuracy of the model is further improved by adjusting the relevant parameters in the algorithm,and the classification results of the model are visually analyzed,and the superiority of the model is compared with the current mainstream algorithms.it realizes the classification model of poor students’ identification and academic early warning,provides referential suggestions for teaching work,and promotes the further development of digitalization and information construction in colleges and universities.