基于BP神经网络的人口预测  

Population prediction based on BP artificial neural network

作　　者：贾楠[1] 胡红萍[1] 白艳萍[1] 

机构地区：[1]中北大学理学院,山西太原030051

出　　处：《山东理工大学学报（自然科学版）》2011年第3期22-24,共3页Journal of Shandong University of Technology:Natural Science Edition

基　　金：国家自然科学基金资助项目(60876077);山西省自然科学基金资助项目(2009011018-3)

摘　　要：利用BP神经网络建立三层神经网络结构模型,以全国人口的出生率,死亡率,老年抚养率等5个指标作为输入样本,对全国总体人口进行预测.隐含层采用多种方法使人为因素降到最低,同时文中采用自适应学习速率和附加动量法相结合的方法,对网络进行训练和学习.将BP预测人口和其他数学方法做比较,预测结果显示该方法是可行的.A three-layer neural network structure model was built using BP neural network model.Five indexes such as the national birth rate,mortality,elderly raising rate were used to predict the whole country population.The hidden layer used a variety of methods to minimize the man-made factors,and this paper used self-adaptive learning speed and additional momentum to train and learn the network.Finally the BP predicted the population,which is compared with other mathematical methods.The resultes showed that this method is feasible to predict population.

关 键 词：BP神经网络 人口预测 神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]