基于误差反向传播算法神经网络模型的网络式以问题为基础学习效果评价方法的研究  

Learning effects evaluation of the WPBL based on BP neural networks model

作　　者：蔡文泳[1] 王心旺[2] 

机构地区：[1]广州医科大学2010级流行病与卫生统计学专业,510182 [2]广州医科大学继续教育学院

出　　处：《中华医学教育杂志》2014年第1期153-156,共4页Chinese Journal of Medical Education

基　　金：广东省广州市教育科学十二五规划课题（11A002）

摘　　要：目的研究误差反向传播算法（ error back propagation algorithm,BP）神经网络模型在网络式以问题为基础学习（ WPBL）效果评价中的适用性。方法基于调查问卷收集的资料,利用测试样本构建BP神经网络模型,并通过所构建的模型对训练样本进行检测。结果构建模型时测试样本的网络输出分数与综合评分之间的平均误差小于规定误差（E=0.000031592〈0.0001）。对训练样本进行检测显示,网络输出分数与实际综合评分之间的误差极小。结论 BP神经网络模型能够准确、快速地对WPBL效果进行评价。Objective To study the applicability of BP neural networks model in the Learning effects evaluation of the WPBL. Methods In this study, based on the data collected through questionnaires, we built the BP neural network model according the test samples and measured training samples according the model. Results The average error between the network output scores and the composite scores when using the test samples to build the model is less than the specified error （ E=0. 000031592〈0. 0001）. For the training samples, the average error between the network output scores and the composite scores very small. Conclusions BP neural network model can evaluate the WPBL learning effects accurately and quickly.

关 键 词：误差反向传播算法 神经网络模型 网络式以问题为基础学习 学习效果评价 

分 类 号：R-4[医药卫生]