基于神经网络的高校管理信息化水平评价模型研究

Research on Evaluation Model of College Management Informatization Level Based on Neural Network

本研究采用频度分析法和领域专家意见相结合,借鉴企业信息化指标的研究结果构建了吉林省高校信息化指标体系;采用了T-S模糊神经网络方法,利用Matlab软件分析,建立高校管理信息化水平评价模型。专家打分200组问卷调查,其中160组结果作为训练样本,40组数据作为测试样本。根据训练好的模型对样本进行测试,训练数据平均误差0.0027,测试数据平均误差0.102。本文研究的T-S模糊神经网络模型应用于预测高校管理信息化评价中可以消除一定的主观因素,得到相对客观的评价结果。

Combined frequency analysis method with expert scoring method, this study established Jilinσs college informatization index system by taking the study results of enterprise index system as reference.Using T-S fuzzy neural network approach analyzed by Matlab software, the evaluation model of college management informatization level was established. 200 questionnaires were scored by expert group, 160 sets of data as the training sample, 40 sets of data as the test samples. According to the samples testing of trained model, the average error of training data was 0.0027, the average error oftest data was 0.102. The T-S fuzzy neural network model applied in the prediction and evaluation of college management information eliminated some subjective factors and obtained the relatively objective evaluation results.