基于模糊神经网络的网络成瘾预测

综合运用模糊数学和神经网络知识构建一个模糊神经网络模型,用以预测网络成瘾。确定了适宜的判别指标和分级标准,对评价论域进行模糊处理;建立各指标对不同论域等级隶属度的计算模型;以实际网络使用者为样本,应用改进的BP算法训练网络模型,并对6个受验样本进行成瘾判别以验证模型的准确性。该方法是对已有的单一指标判别法和用模糊数学对多个指标判别方法的改进。实验证明,改进的BP神经网络方法能够快速、准确、有效地识别网络成瘾模式。

Forecasting Pattern of Network Addiction Based on Fuzzy Neural Network

In order to forecast pattern of network addiction,a fuzzy neural network model using fuzzy mathematics and neural network is set up.Proper judgment indexes,classification standards and fuzzy treatment to assessment sets are performed.The calculation method for subordinate degrees to different set grades of each index is built.Trained based on improved BP algorithm with training samples,the model is validated by forecasting addiction properties of six users.This method is an improvement of single index judgment and multi-index judgment based on fuzzy mathematics.The experiment results show that the approach could recognize the pattern of addiction rapidly,accurately and effectively.