人工神经网络在基础课实验教学质量评价体系中的应用  

Application of artificial neural network in teaching quality evaluation system of basic experiment

作　　者：高艳萍[1] 周敏[1] 郭显久[1] 何东钢[1] 

机构地区：[1]大连水产学院信息工程学院,辽宁大连116023

出　　处：《辽宁师范大学学报（自然科学版）》2008年第4期433-435,共3页Journal of Liaoning Normal University:Natural Science Edition

基　　金：辽宁省教育厅攻关项目(05L090)

摘　　要：采用神经网络对基础课实验教学质量进行评价,解决了非线性模型的不确定性.传统的BP网络在训练时易陷入局部极小点,从而导致训练时间长、收敛速度慢.采用优化BP算法能快速达到目标误差,提高网络的收敛速度.网络训练的结果表明:优化BP算法的收敛速度快,精度高,在高等教育教学质量评估领域中具有广阔的应用前景.The project is to evaluate teaching quality of basic experiment with neural Network. Since teaching quality evaluation system is a nonlinear system difficult to deal with,the artificial neural network present an optimization tool for it. However, the training of neural network by conventional back-propagation（BP） method has intrinsic vulnerable weakness with slow convergence and local minima, BP optimized algorithm not only possesses the advantages of artificial neural network, but also offsets the disadvantages caused by the BP neural network. The experimental results by MATLAB show that this method is effective and more accurate than BP neural network, and it can be applied in teaching quality of higher education.

关 键 词：反向传播(BP) 教学质量 评价体系 神经网络 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]