【摘要】 选取影响手机网民规模的30个变量,借助于主成分分析方法进行降维,以降维后的5个主成分变量作为手机网民规模预测模型的输入层变量,采用BP和RBF神经网络分别对手机网民规模进行分析和预测。研究结果显示,采用BP神经网络,预测2016年12月与2017年6月的手机网民规模分别为69046(万人)和72359(万人);采用RBF神经网络,预测2016年12月与2017年6月的手机网民规模分别为68702(万人)和71972(万人)。更多还原

【Abstract】 Select the 30 variables affecting mobile phone netizens’ scale, decrease the dimension to 5 by principal component analysis method, and the 5 principal component variables are used as the input layer variables of mobile phone netizens’ scale prediction model. BP and RBF neural network theories are used to analyze and forecast the scale of mobile phone netizens. The research results show that, according to the BP neural network, the forecast scale of mobile phone netizens are 69046(million)and 72359(million) respectively in December 2016 and June 2017; and according to the RBF neural network, the forecast scale are 68702(million) and 71972(million) respectively in December 2016 and June 2017.更多还原