基于EMD.ARXG模型的网络舆情预测研究

Research on network public opinion prediction based on EMD.ARXG model

针对复杂的网络舆情数据,传统的模型预测已经无法对大数据背景下的舆论趋势进行有效的预测,因此,提出一种基于EMD.ARXG(经验模态分解⁃自回归)改进的组合模型来应对复杂的网络舆情预测,该模型弥补了单一预测算法的缺陷,提高了预测模型的准确性。以“韩国萨德”事件和“全国两会”事件作为舆情热点对其进行预测实验,引入WNN(小波神经网络)与EMD⁃BPNN(BP神经网络)进行舆情预测,并与EMD.ARXG模型进行实验对比,实验结果证明,EMD.ARXG模型具有较好的预估准确度。

Since the network public opinion data is complex,traditional model prediction has been unable to effectively predict the trend of public opinion under the background of large data.Therefore,an improved combination model based on EMD.ARXG(empirical mode decomposition⁃autoregression)is proposed to deal with complex network public opinion prediction.This model makes up for the shortcomings of a single prediction algorithm and improves the accuracy of the prediction model.Taking the events of"THAAD in Korea"and"NPC&CPPCC"as the hotspots of public opinion,WNN(wavelet neural network)and EMD⁃BPNN(BP neural network)is introduced to predict public opinion,and compares them with the EMD.ARXG model.The experimental results show that the EMD.ARXG model has better prediction accuracy.