摘要
采用三层结构的反向传播网络(back-propagation network,简称BP网络),对山东省国内游客量建立了网络预测模型.利用实际数据进行模拟,并将模拟结果与采用相同数据源的灰色系统模型、多元回归模型、指数模型的模拟结果进行比较,发现用BP网络模型的模拟结果与实际情况最为逼近.同时,利用该模型对未来几年山东省的国内游客量进行了分析预测,预期到2008年山东省的国内游客量将达到12 577万人次.
Network forecasting model of domestic tourists in Shandong by three-layer Back-propagation network is established. It is found that the simulated result of B-P network model is the most approached practical situation after making comparison with the simulated results of other models such as grey system, multi-regression, exponential regression. And the amount of domestic tourists in Shandong in future is forecasted with this model. The prediction amount will be 125770000 in the year of 2008.
出处
《中北大学学报(自然科学版)》
EI
CAS
2007年第5期426-429,共4页
Journal of North University of China(Natural Science Edition)
关键词
人工神经网络
B—P算法
游客量
预测
artificial neural networks
B-P algorithm
domestic tourism scale
forecast