融合时间序列和协同过滤的学生成绩预测方法  

Student Grade Prediction Method Integrating Time Series and Collaborative Filtering

作　　者：单春宇 张怡文[2] 张婷 吉宇洁 褚俊 董云春 SHAN Chunyu;ZHANG Yiwen;ZHANG Ting;JIE Yujie;CHU Jun;DONG Yunchun(School of Electronics and Information Engineering,Anhui Jianzhu University,230601,Hefei,Anhui,China;College of Information Engineering,Anhui Xinhua University,230088,Hefei,Anhui,China;School of Mathematics and Physics,Anhui Jianzhu University,230601,,Hefei,Anhui,China)

机构地区：[1]安徽建筑大学电子与信息工程学院,安徽合肥230601 [2]安徽新华学院信息工程学院,安徽合肥230088 [3]安徽建筑大学数理学院,安徽合肥230601

出　　处：《淮北师范大学学报（自然科学版）》2022年第3期69-74,共6页Journal of Huaibei Normal University：Natural Sciences

基　　金：安徽省教育厅自然科学重点项目(KJ2020A0784);安徽省省级大学生创新创业训练计划项目(AH202012216092)。

摘　　要：随着高校招生规模的不断扩大,教师对学生学习掌握情况的全面了解变得更加困难,为更好地掌握学生的学习情况,降低学生的不及格率,学生成绩预测显得尤为重要.目前采用协同过滤方法对学生成绩预测,由于课程之间具有较强的时间序列特征,学生对课程的兴趣也会随着时间的推移发生动态变化,导致考试成绩预测准确率较低.针对该问题,提出融合时间序列和协同过滤的学生成绩预测方法.实验结果表明,基于时间序列和协同过滤的学生成绩预测方法对学生成绩预测的RMSE、MAE误差有所降低.With the continuous expansion of the enrollment scale of colleges and universities in our country,it is increasingly difficult for teachers to grasp the students′learning situation.To better grasp the students′learning situation and reduce the failure rate of students,it is particularly important to predict students′grades.At present,the collaborative filtering method is mainly used to predict students′grades,but there are strong time-series characteristics among courses,and the students′interest in the courses will change over time,resulting in low accuracy of test score prediction.In response to this problem,a student grade prediction method combining series and collaborative filtering is proposed.The results show that the error of RMSE and MAE prediction of the students′performance based on time series and collaborative filtering is reduced.

关 键 词：时间序列 协同过滤 学业成绩预测 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]