混合神经网络的高校网络舆情预测模型研究  

Research on College Internet Public Opinion Prediction based on Hybrid Artificial Neural Network

作　　者：武燕[1] 黄颜[1] 吴维[2] 李孟山[2] 

机构地区：[1]赣南师范大学数学与计算机科学学院,江西赣州341000 [2]赣南师范大学物理与电子信息学院,江西赣州341000

出　　处：《赣南师范学院学报》2016年第6期23-26,共4页Journal of Gannan Teachers' College(Social Science(2))

基　　金：国家自然科学基金项目(51663001);江西省教育厅科学技术研究项目(GJJ150983;GJJ151012);校级教改项目(GSJG-2015)

摘　　要：为了对高校大学生群体性事件网络舆情进行准确预测并作正确引导,提出一种基于改进粒子群算法的混合神经网络(HANN)的高校网络舆情的发展趋势预测模型.HANN首先采用自适应调整策略和混沌理论对粒子群算法进行改进得到改进粒子群算法(CSA-PSO),再通过CSA-PSO算法训练径向基人工神经网络(RBF ANN)得到;RBF ANN的结点个数通过试探法确定.通过实例测试和与其它模型比较实验,结果表明所提出的HANN方法具有较高的预测精确,综合性能较好.In order to accurately predict and guide internet public opinion of college students group event, in this paper, a hybrid neural network （HANN） model based on improved particle swarm optimization algorithm is proposed to predict the development trend of network public opinion in universities. The Model forecasts the development trend of internet public opinion in Colleges and Universities. Firstly, the particle swarm optimization algorithm is improved by the adaptive adjustment strategy and the chaos theory, and the improved particle swarm optimization algorithm （CSA-PSO） is proposed. CSA-PSO algorithm is used to train RBF ANN, and the number of nodes in the hidden layer is determined by the heuristic method. Finally, HANN is proposed and used to predict the samples. Compared with other models, the experimental results show that the proposed HANN has a high degree of accuracy and has good performance.

关 键 词：网络舆情 群体性事件 神经网络 混沌理论 自适应调整策略 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP393[自动化与计算机技术—控制科学与工程]