基于组合赋权和BP神经网络的高校学生综合素质评价研究  

Research on Comprehensive Quality Evaluation of College Students Based on Combined Weight and BP Neural Network

作　　者：尤游[1] 张林静[2] You You;Zhang Linjing(Anhui Technical College of Mechanical and Electrical Engineering,Wuhu 241000,China;Anhui Vocational College of Police Officers,Hefei 230031,China)

机构地区：[1]安徽机电职业技术学院公共基础教学部,安徽芜湖241000 [2]安徽警官职业学院信息管理系,安徽合肥230031

出　　处：《安徽电子信息职业技术学院学报》2021年第5期5-9,共5页Journal of Anhui Vocational College of Electronics & Information Technology

基　　金：安徽省2020年省级质量工程教学研究项目(2020jyxm0310);安徽省2020年高校自然科学研究项目(KJ2020A1058)。

摘　　要：考虑到主客观赋权的弊端,以高校学生综合素质评价为例,提出了一种基于最小相对信息熵原理的组合赋权方法,建立了综合职业能力视角下的评价模型。为了提高综合评价的便捷性,以单项指标为输入样本,评价成绩为输出样本构建BP神经网络。仿真结果表明,BP神经网络的预测平均相对误差为0.61%,最大相对误差为1.57%,说明该模型在合理的范围内能快速给出评价结果,避免了繁琐的计算过程,具有一定的应用前景。Considering the disadvantages of subjective and objective weight,taking the comprehensive quality evaluation of college students as an example,a combined weight method based on minimum relative information entropy principle is proposed,and an evaluation model from the perspective of comprehensive occupation ability is established.In order to improve the convenience of comprehensive evaluation,BP neural network is constructed with single index as input sample and evaluation result as output sample.Simulation results show that the average relative error of BP neural network prediction is 0.61%,and the maximum relative error is 1.57%.It shows that the model can quickly give evaluation results within a reasonable range,avoiding tedious calculation process,and has certain application prospect.

关 键 词：组合赋权 BP神经网络 最小相对信息熵原理 综合职业能力 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]