全国私人车辆拥有量的BP神经网络模型预测与分析——基于附加动量与自适应学习速率相结合的BP方法  

Prediction and Analysis of Private Car Ownership in China Based on the BP Neural Network:Based on Additional Momentum and Adaptive Learning Rate of BP Method

作　　者：陈正[1] 丁姝[1] 王俊林[2] 

机构地区：[1]西安财经学院统计学院,陕西西安710100 [2]天津财经大学数学经济研究中心,天津300222

出　　处：《西安财经学院学报》2015年第6期98-102,共5页Journal of Xi’an University of Finance & Economics

摘　　要：私人车辆拥有量的变化趋势与社会经济发展密切相关,研究与预测私人车辆拥有量变动对了解社会经济发展具有重要的意义。文章以1985—2013年全国私人车辆拥有量的发展情况为研究对象,利用附加动量法与自适应学习速率相结合的BP神经网络模型对全国私人车辆拥有量进行研究与预测,结果表明利用该模型对私人车辆拥有量进行预测的精度较高、效果好。T he change trend of private car ow nership is closely related to the social and economic develop‐ment ,so accurately predict the private car ow nership has very important significance .In this paper ,from 1985 to 2013 ,the national development of private car ow nership situation as the research object ,using the combination of additional momentum method and adaptive learning rate of BP neural network model predict and analyze the private car ow nership .T he prediction results show that the additional momentum and adap‐tive learning rate combined with BP neural network model to predict the national private car ownership ,has high precision and good prediction effect .

关 键 词：私人车辆拥有量 BP神经网络 附加动量法 自适应学习速率法 预测模型 

分 类 号：F224[经济管理—国民经济] TP183[自动化与计算机技术—控制理论与控制工程]