灰色神经网络在云南省旅游收入中的预测研究

Forecasting Tourism Income of Yunnan Province by Gray Neural Network

随着生活水平的不断提高,旅游业迎来了蓬勃的发展,其中国内旅游收入是衡量旅游业发展的重要依据。为了正确预测旅游收入,本文以灰色神经网络模型预测值与真实值的误差平方和最小为原则确定GM(1,1)灰色模型与BP神经网络的组合权重建立灰色神经网络模型,以讨论云南省旅游收入预测问题。选取1997-2011年云南省旅游收入实际数据为建模数据,以2012-2016年的实际数据为检验数据,对比分析GM(1,1)、BP神经网络和灰色神经网络的预测精度。其结果表明:灰色神经网络的预测精度最高,其预测误差均方差为0.24,小于GM(1,1)和BP神经网络预测误差的均方差。因此,基于灰色神经网络的组合预测模型具有较高的预测精度。

With the continuous improvement of living standards, tourism ushers in the opportunity to flourish where domestic tourism income is an important measure of tourism development. In order to forecast the domestic tourism revenue correctly, a gray neural network model for Yunnan province is established employing the combination weights of GM（1,1） modle and BP Neural network which are determined by the minimal the square sum of the error of prediction. Selecting the actual data of tourism income of Yunnan Province in1997-2011 year as the built samples, and the data 2012-2016 as testing samples, we study the prediction accuracy of GM（1,1）, BP neural Network and grey neural network. The results show that the grey Neural network has the highest prediction accuracy with the mean square error is 0.72, which is less than the GM（1,1） and BP neural networks prediction error variance. Therefore, the grey neural network has a high prediction precision which can be used for forecasting the tourism income.