基于深度学习的远程教育学生流失预测模型的建立与评估  

Establishment and Evaluation of Distance Education Student Churn Prediction Model Based on Deep Learning

作　　者：纪娟 JI Juan(Sichuan Open University,Chengdu 610073,China;Educational Information Management and Information System Research Center in the Open University of China,Beijing 100039,China)

机构地区：[1]四川开放大学,成都610073 [2]国家开放大学教育信息管理与信息系统研究中心,北京100039

出　　处：《北京工业职业技术学院学报》2022年第3期21-26,共6页Journal of Beijing Polytechnic College

摘　　要：从在线学习的完整学业流程来看,以往基于在线学习平台的课程学习行为数据的学生流失预测,在预判学生流失因素方面存在片面性和不完整性。针对这一问题,在分析远程教育学生核心业务的基础上,通过量化学生在核心业务上的活跃度,加入可以影响学生流失的关键基础属性,使用深度学习算法,建立基于深度学习的远程教育学生流失预测模型。评估模型表明:构建的学生流失预测模型达到了理想的预测效果,可为全面和深入地分析远程教育中学习者的流失因素提供科学依据。From the perspective of the complete academic process of online learning,the previous prediction of student churn based on the course learning behavior data of online learning platform is one-sided and incomplete in predicting the factors of student churn.To solve this problem,on the basis of analyzing the core business of distance education students,by quantifying the activity of students in the core business,adding the key basic attributes that can affect the churn of students,and using the deep learning algorithm,the author establishes a prediction model of distance education student churn based on deep learning.The evaluation model shows that the constructed student churn prediction model has achieved the ideal prediction effect,which provides a scientific basis for a comprehensive and in-depth analysis of the churn factors of learners in distance education.

关 键 词：流失预测 深度学习 神经网络 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]