基于多元线性回归的数学建模成绩预测研究

Research on mathematical modeling achievements predictionbased on multiple linear regression

利用机器学习中的多元线性回归方法建立了数学建模成绩的预测模型。首先,选择课程并进行数据清洗,确定数据集。其次,利用相关性分析和方差膨胀因子检验课程成绩间的相关性和多重共线性,通过逐步回归的方式确定数学建模成绩的影响因素和回归模型。最后,利用交叉验证,在不同的训练集上训练模型,利用均方误差和平均相对误差检验模型的预测准确性。结果表明,采用多元线性回归预测的数学建模成绩与实际成绩相近,预测模型有效。

A mathematical modeling achievements prediction model was established using the multiple linear regression method in machine learning.Firstly,the dataset was determined by selecting prerequisite courses,data cleaning,etc.Secondly,correlation analysis and variance inflation factor were used to test the correlation and multicollinearity between course achievements.The stepwise regression method was used to determine the influencing factors and regression model of mathematical modeling achievement.Finally,the cross validation method was used to train the model on different training sets,and the prediction accuracy of the model was tested using mean square error and mean relative error.The experimental results show that the mathematical modeling achievements predicted by multiple linear regression are similar to the actual values,the effectiveness of the model is verified.