基于SVDD算法的学业预警监测模型构建与研究

Construction and Research of Early Academic Warning Monitoring Model based on SVDD Algorithm

学生在在线学习平台上的学习表现数据类型较多,不能从单一角度对学生学习状况进行评价监测,需要综合多维度评价学生的学习状况,从而给出相对全面客观的学习表现评价。建立基于SVDD算法的监测模型,将直播观看时长、作业完成状况、课堂参与互动次数、出勤率等多维数据作为考量指标,对所有数据进行PCA分析,通过SVDD算法进行学习行为和效果异常监测,从而实现表现异常学生的学业预警。

There are many types of students'learning performance data on online learning platforms,and it is not possible to evaluate and monitor students'learning status from a single perspective.It is necessary to comprehensively evaluate students'learning status from multiple dimensions to give a relatively comprehensive and objective learning performance evaluation.The paper establishes a monitoring model based on the SVDD algorithm,which takes multidimensional data such as live viewing time,homework completion status,classroom participation and interactions,and attendance rate as consideration indicators,and performs PCA analysis on all data,and monitors learning behavior and effect abnormalities through SVDD algorithm so as to realize the academic warning of abnormal students.