摘要
针对高校当前在贫困生评定上存在效率低、不透明、不客观等问题,提出了学生贫困评定模型用以缓解上述问题。该模型首先假设高校学生评定数据与贫困类别之间存在特定的函数关系;其次,在贫困数据中提取了反映贫困学生的真实且易于获取的2大类10个评价指标并对其进行了度量;再次,利用LVQ(Learning Vector Quantization)神经网络表示学生数据与贫困类别之间的特定函数关系;最终,实验结果表明了假设的合理性与模型的正确性且模型可以达到95.7%的准确度并以3.6%精确度优于普遍使用的决策树方法。
Aiming at solving the no-efficient and nonobjective problem existing in the evaluation of the needy student in colleges and universities,this paper proposes the model for evaluating the poverty of the student to address such problem.First,the proposed model assumes that there is a special functional relationship between the evaluation data of student in the university and the poverty category;then,extracts and measures two major categories (ten items) evaluation indexes that can be easily got from the data of needy student;then,uses the LVQ (Learning Vector Quantization) neural network to represent the special functional relationship between the data of student and the category.Finally,the experimental result demonstrates the reasonableness of the assumption and the correctness of the model.The result also shows the proposed model can achieve 95.7% accuracy and excels the decision tree that commonly used by higher accuracy of 3.6%.
作者
刘凯
黄胜
李君科
LIU Kai;HUANG Sheng;LI Jun-ke(School of Educational Science,Qiannan Normal University for Nationalities,Duyun 558000,Guizhou,China)
出处
《黔南民族师范学院学报》
2019年第4期70-73,共4页
Journal of Qiannan Normal University for Nationalities
基金
贵州省科学技术基金重点项目(黔科合LH字[2014]7422)阶段性成果
关键词
贫困生评定
评定指标
指标度量
评定建模
LVQ神经网络
evaluation of the needy student
evaluation index
index measurement
evaluation model
LVQ neural network
