基于神经网络的我国汽车保有量建模与预测

利用改进 BP神经网络对我国 2 0 0 0年、2 0 0 5年、2 0 10年的汽车保有量进行预测 ,并与灰色预测的结果进行对比 ,结果表明 ,由于神经网络具有很强的学习与泛化能力 ,在处理具有一定程度不确定性的非线性系统的建模与预测方面有较好的应用价值。

This paper forecasts automobile\|owned of China in 2000 and in 2005 and in 2010 based on BP neural network, and compares the results with those of gray forecasting. It has shown BP network is very available with respect to modeling and forecasting of nonlinear system which has uncertainties because the network has reinforcement learning properties and mapping capabilities.