摘要
大学生的学习成绩与其学习状态和习惯有正相关性。教师工作手册中记录的考勤、答问与作业信息反映了学生的学习状态,智慧课堂中的随堂提问、课后作业、座位偏好等信息进一步反应出学生的行为习惯。充分利用上述数据进行期末成绩预测并向学生反馈学业警示和鼓励信息,将对教学起到积极作用。设计了PSO-BP神经网络预测模型来进行学生行为数据挖掘,筛选了具有代表性的数据作为神经网络的输入,选择课程成绩作为神经网络的输出,成绩预测误差为12%,为提高教学质量提供了新的思路。
College students’academic performance is positively correlated with their learning status and habits.The attendance,Q&A,and homework information recorded in the teacher’s work manual reflects the student’s learning status.In the smart classroom,information such as in-class questions,homework,seat preference,etc.further reflects the behavior of students.Making full use of the above-mentioned data to predict the final grade and feedback the academic warnings and encouraging information to students will play a positive role in teaching.The PSO-BP neural network prediction model is designed to conduct student behavior data mining,representative data is selected as the input of the neural network,and the course score is selected as the output of the neural network,and the score prediction error is 12%,which provides a new idea for improving the teaching quality.
作者
郭涛
魏勇
熊杰
Guo Tao;Wei Yong;Xiong Jie(Electronics&Information School,Yangtze University,Jingzhou,Hubei 434023,China;Huanggang Normal College)
出处
《计算机时代》
2021年第3期52-56,共5页
Computer Era
基金
教育部科技发展中心高校产学研创新基金(2018A03009)。
关键词
PSO算法
BP算法
神经网络
成绩预测
智慧课堂
数据挖掘
PSO algorithm
BP algorithm
neural network
performance prediction
smart classroom
data mining
