基于t检验和随机森林算法的高等数学成绩主控因素分析

Analysis of Main Control Factors of Advanced Mathematics Performance Based on t-Test and Random Forest Algorithm

摘要: 围绕不同专业学生学习高等数学成绩主控因素筛选问题，构建了两个专业236名学生的高等数学学习过程指标和结课成绩数据集。运用描述性统计分析方法刻画成绩分布特征，并借助t检验判断专业间学习数据和成绩差异的显著性。使用随机森林模型，对影响学生高数成绩的进行分析，发现课上习题作答率和课后习题得分率对结课成绩的影响更加显著，到课率对结课成绩影响最不显著。基于数据分析结果，针对性提出了面向所学专业进行差异化教学以及提升课堂互动、课后学习质量两大类教学改进措施。

Abstract: Regarding the screening of the dominant factors influencing the achievements of students from different majors in advanced mathematics, a dataset of the learning process indicators and final course scores of 236 students from two majors in advanced mathematics was constructed. The descriptive statistical analysis method was used to depict the characteristics of the score distribution, and with the help of the t-test, the significance of the differences in learning data and scores among different majors was judged. By using the Random Forest model to analyze the factors affecting students’ scores in advanced mathematics, it was found that the response rate of in-class exercises and the score rate of after-class exercises had a more significant impact on the final course scores, while the attendance rate had the least significant impact on the final course scores. Based on the results of data analysis, two major types of teaching improvement measures were put forward in a targeted manner, including differentiated teaching according to the majors studied and improving the quality of classroom interaction and after-class learning.