无线电工程

针对基于BP神经网络的课堂教学质量评价存在预测误差较大的不足,为提高评价的准确性和预测性,根据不同评价主体的特点,提出了基于层次分析法和问卷调查法相结合来选取评价指标体系,作为BP神经网络的输入;进而运用遗传算法选择最优个体的原理,优化BP神经网络初始参数来建立课堂教学质量的评价模型。通过对样本数据的训练与测试,表明该模型较大地提高了课堂教学质量评价的精度,在教学评价中具有较好的应用前景。

The classroom teaching quality evaluation based on BP neural network has the shortage of large prodiction error. In order to improve assessment accuracy and predictability,according to characteristic of different evaluation subject,an improved assessment index system as the input of BP neural network is applied in the paper based on analytic hierarchy process and questionnaire survey method; and with the optimized BP neural network initial parameters by the principle of selecting optimal individual from genetic algorithms,an improved classroom teaching quality evaluation model is established.The simulation result shows that it can greatly improve assessment accuracy by neural network training and testing from the sample data,and has a good application prospects for teaching quality evaluation.