With increasing difficulty on preliminary test of postgraduate entrance examination as well as numbers of factors affecting results of the examination,taking college students majoring in architectural design of Anhui University of architecture as the research ob-ject,this study uses grades at school and the preliminary test grades of the undergraduate graduates as the sample data,through the com-parison of three experimental modeling tests of Logistic regression classification algorithm,SVM algorithm and KNN algorithm,to find cor-responding change rules and improve the correlation between score variables from preliminary tests of postgraduate entrance examination,so as to obtain the prediction results.From the perspective of average prediction error,the prediction method of logistic regression classifi-cation algorithm has high adaptability and stability,and the accuracy is more suitable for college students majoring in architectural design without higher mathematics in preliminary tests.It is concluded that special attention should be paid to strengthening the study of special-ized courses and curriculum design,while paying attention to the learning attitude towards politics and English.The conclusion helps them predict their academic development trend and provides data support for career planning.
关键词
考研初试成绩预测/Logistic算法/SVM算法/KNN算法/建筑类学生
Key words
prediction of postgraduate entrance examination results/logistic algorithm/SVM algorithm/KNN algorithm/architecture students
维普
万方数据