改进GA-BP神经网络在高校教学质量评价中的应用  

Application of improved GA-BP neural network in teaching quality evaluation

作　　者：岳琪[1] 温新 YUE Qi;WEN Xin(College of Information and Computer Engineering,Northeast Forestry University,Harbin 150040,China)

机构地区：[1]东北林业大学信息与计算机工程学院

出　　处：《黑龙江大学自然科学学报》2019年第3期353-358,共6页Journal of Natural Science of Heilongjiang University

基　　金：国家自然科学基金资助项目(61702091);黑龙江省教育厅规划课题(GJC1316036)

摘　　要：为了提高高校教学质量评价的有效性和准确性,提出了一种基于混合智能优化算法的教学质量评价模型。引入熵值法客观地确定教学质量评价体系的指标权重及初始评价结果,利用自适应变异的遗传算法(Genetic algorithm)优化BP神经网络(Back propagation neural network )的参数,建立教学质量评价模型。实验结果表明,与BPNN(Back propagation neural network)、GA-BPNN(Genetic algorithm-back propagation neural network)模型相比,预测精度分别提高15.04%和 5.41%,收敛速度明显提高,说明基于自适应变异的GA-BPNN教学质量评价模型能够及时有效地完成教学质量评价。In order to improve the validity and accuracy of teaching quality evaluation, a teaching quality evaluation model is proposed based on hybrid intelligent optimization algorithm. The entropy method is introduced to determine the index weights of the teaching quality evaluation system and the initial evaluation result objectively. Adaptive mutation based GA (Genetic algorithm) is used to optimize the parameters of BP neural network (BPNN) to establish the teaching quality evaluation model. Compared with BP neural network and GA-BPNN (Genetic algorithm-back propagation neural network) models, the prediction accuracy is improved by 15.04% and 5.41%, respectively. The convergence rate is also improved. It shows that GA-BPNN teaching quality evaluation model based on adaptive mutation can complete the teaching quality evaluation timely and effectively.

关 键 词：熵值法 遗传算法 自适应变异概率 BP神经网络 教学质量评价模型 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]