摘要
基于M2B学习支持系统提供的数据,针对课程学习活动日志数据特性,提出了一种用双向门控单元网络(Bi-GRU)预测学生学习成绩的方法.将已有的递归神经网络(RNN)方法作为基准线,通过实验比较证实了Bi-GRU用于成绩预测的有效性.
Based on the data provided by the M2 Blearning support system,this paper proposes a method for predicting students′academic performance by using Bidirectional Gated Recurrent Unit network(Bi-GRU)targeting at the characteristics of course learning activity log data.The existing Recurrent Neural Network(RNN)method was taken as the reference line,and the effectiveness of Bi-GRU for performance prediction was verified by the experimental comparison.
作者
张昊
ZHANG Hao(College of Science,Tieling Normal College,Tieling Liaoning 112008)
出处
《辽宁师专学报（自然科学版）》
2019年第3期94-96,共3页
Journal of Liaoning Normal College(Natural Science Edition)
