基于Simpson公式的灰色神经网络在GDP预测中的应用  

Application of Simpson Formula-based Grey Neural Network in GDP Forecasting

作　　者：何刚[1] 吴文青[2] 夏杰 He Gang;Wu Wenqing;Xia Jie(College of Computer Science and Technology,Southwest University of Science and Technology,Mianyang Sichuan 621010,China;School of Science,Southwest University of Science and Technology,Mianyang Sichuan 621010,China)

机构地区：[1]西南科技大学计算机科学与技术学院,四川绵阳621010 [2]西南科技大学理学院,四川绵阳621010

出　　处：《统计与决策》2020年第2期43-47,共5页Statistics & Decision

基　　金：西南科技大学博士基金资助项目(16zx7108)。

摘　　要：文章基于Simpson公式改进的GM(1,1)灰色系统和神经网络组合模型对国内生产总值进行预测研究。首先,利用Simpson积分公式对GM(1,1)灰色系统的背景值进行改进。其次,通过相关性分析确定财政收入、财政支出、全社会固定资产投资、进出口差额、国家税收收入和社会消费零售总额6个因素为GDP的主要影响因素。接着,将灰色系统的预测值和影响GDP总量的6个因素同时作为BP神经网络的输入构建串联型灰色神经网络预测模型。对比分析GM(1,1)、Simpson公式改进的GM(1,1)、Simpson公式改进的灰色神经网络模型的计算结果,可明显看出基于Simpson公式改进的灰色神经网络预测精度最高。This paper makes a forecast study on gross domestic product(GDP)by using the improved GM(1,1)grey system and the neural network combination model based on Simpson formula.Firstly,Simpson integral formula is employed to improve the background value of GM(1,1)gray system.Secondly,through correlation analysis,the main influencing factors of GDP are identified as fiscal revenue,fiscal expenditure,investment in fixed assets of the whole society,balance of import and export,national tax revenue and total retail consumption.Finally,the predicted value of the gray system and the six factors that affect the total GDP are simultaneously used as the input of the BP neural network to construct the serial gray neural network prediction model.By comparing and analyzing the calculation results of GM(1,1),the GM(1,1)improved by Simpson formula,and the grey neural network model improved by Simpson formula,it is obviously displayed that the improved grey neural network based on Simpson formula has the highest prediction accuracy.

关 键 词：GM(1 1)模型 BP神经网络 SIMPSON公式 GDP 

分 类 号：N941.5[自然科学总论—系统科学]