基于教育数据挖掘的在线学习者学业成绩预测建模研究

发布时间：2019-05-11 15:41

【摘要】：该文采用数据挖掘和机器学习的方法,研究从教育数据中挖掘影响在线学习者学业成绩的因素并构建分类预测模型。首先,通过计算所有单个数据属性和学业成绩类别之间的相关系数和计算所有属性的信息增益率两种方法共同确定学业成绩的影响因素。然后,提出采用集成学习的方法构建集成式学业成绩分类预测模型,并比较多种算法构建的单一分类模型和集成分类模型的性能。最后,进一步采用嵌套集成学习的方法构建在线学习者学业成绩分类预测模型,并对模型的性能进行评估。研究成果可以为在线学习者学业成绩影响因素研究和预测建模研究提供借鉴,也有助于推进在线学习学业预警、学业成绩预测和评价的实践。
[Abstract]:In this paper, data mining and machine learning are used to study the factors that affect the academic performance of online learners from educational data and to construct a classified prediction model. Firstly, the influencing factors of academic achievement are determined by calculating the correlation coefficient between all individual data attributes and academic achievement categories and calculating the information gain rate of all attributes. Then, the integrated learning method is proposed to construct the integrated academic achievement classification prediction model, and the performance of the single classification model and the integrated classification model constructed by various algorithms is compared. Finally, the nested ensemble learning method is used to construct the online learners' academic achievement classification prediction model, and the performance of the model is evaluated. The research results can provide a reference for the research of influencing factors and prediction modeling of online learners' academic performance, and also help to promote the practice of online learning academic early warning, academic achievement prediction and evaluation.
【作者单位】：华中师范大学国家数字化学习工程技术研究中心;贵州财经大学信息学院;
【基金】：国家科技支撑计划课题“学习云服务平台关键技术研究与平台开发”(项目编号:2015BAH33F02)研究成果;
【分类号】：G434

【相似文献】