摘要
建设人口均衡城市,首要在于充分摸清人口底数,准确预测人口增长规模,才能科学制定人口和经济社会发展规划。文章针对以往单一指标预测人口总量精度不高的问题,利用灰色预测所需信息少和神经网络非线性映射能力强的优点,筛选出与人口总量紧密联系的9个主要因素,力图建立多指标灰色PSO-BP神经网络人口预测模型,并对2015年至2030年成都市人口总量进行预测。实证分析表明,该模型预测精度高、泛化能力强,具有较好的实用价值。在论文的最后,作者依据模型预测结果,为成都市建设人口均衡城市提出了较有针对性的建议。
Establishing a city with balanced population, the primary factor is to find out the base of the population completely and to predict the scale of population growth accurately. Then we could make feasible development plan for population and economic society. For the precision of population prediction via single indicator is lacking, this paper takes advantages of the grey prediction which requires less information and the neural network whose nonlinear mapping ability is strong, to screen 9 main factors closely connected with the total population, aim to build multi-indicator gray PSO-BP neural network model of population prediction in order to predict the total population in Chengdu between 2015 and 2030. The empirical analysis shows that the prediction model is accurate and has strong generalization ability and good practical values. At the end of the paper, based on prediction data of the model, some targeted suggestions are presented for constructing Chengdu into a population balanced city by the author.
出处
《技术经济与管理研究》
CSSCI
2014年第6期90-95,共6页
Journal of Technical Economics & Management
基金
成都市哲学社会科学规划项目(ZST12-02
ZST11-06)
