大数据时代高校学籍预警机制的探索与研究
发布时间：2019-07-09 07:03
【摘要】：随着高校教育规模的不断扩大,现有的教学管理模式突现出来的问题越来越多。在大数据时代背景下,采集学生的基本信息、考勤信息、上网信息、图书借阅信息、校园消费信息等特征作为学生标签,对学生特征汇总形成学生画像,对采集信息进行预处理生成学生特征矩阵,通过主成分分析法对学生特征矩阵进行降维处理,最后利用基于距离的聚类分析技术对经过降维后的学生特征数据分类,分离出偏离中心点的学生特征,获得学籍异常学生的信息,从而可以对学籍状态进行动态监测,对学籍状态异常学生进行预警和帮扶指导。
[Abstract]:With the continuous expansion of the scale of university education, there are more and more problems in the existing teaching management mode. Under the background of big data's era, the basic information of students, attendance information, Internet information, book borrowing information, campus consumption information and other features are collected as student labels, the characteristics of students are summarized to form student portraits, the collected information is preprocessed to generate student feature matrix, and the dimension reduction of student feature matrix is carried out by principal component analysis. Finally, the distance-based clustering analysis technology is used to classify the students' characteristic data after dimension reduction, to separate the students' characteristics that deviate from the center point, and to obtain the information of the students with abnormal student status, so that the student status can be monitored dynamically, and the students with abnormal student status can be given early warning and help guidance.
【作者单位】： 北京信息科技大学工程训练中心;
【基金】：北京信息科技大学教学改革项目(2015JGYB44)研究成果
【分类号】：G647.3
，
本文编号：2511957
[Abstract]:With the continuous expansion of the scale of university education, there are more and more problems in the existing teaching management mode. Under the background of big data's era, the basic information of students, attendance information, Internet information, book borrowing information, campus consumption information and other features are collected as student labels, the characteristics of students are summarized to form student portraits, the collected information is preprocessed to generate student feature matrix, and the dimension reduction of student feature matrix is carried out by principal component analysis. Finally, the distance-based clustering analysis technology is used to classify the students' characteristic data after dimension reduction, to separate the students' characteristics that deviate from the center point, and to obtain the information of the students with abnormal student status, so that the student status can be monitored dynamically, and the students with abnormal student status can be given early warning and help guidance.
【作者单位】： 北京信息科技大学工程训练中心;
【基金】：北京信息科技大学教学改革项目(2015JGYB44)研究成果
【分类号】：G647.3
，
本文编号：2511957
本文链接：https://www.wllwen.com/jiaoyulunwen/shifanjiaoyulunwen/2511957.html
最近更新
教材专著
