摘要
人口问题是一个复杂的社会问题,准确地预测,对于国家社会发展计划特别是制定生育政策有重要的意义。本文探讨了基于遗传算法改进的BP神经网络模型的原理、构成以及实现方式,该改进模型可以有效提高BP网络模型的性能。同时将此模型用于我国人口数量的预测,选取我国1982-2010年的人口数据,检验结果平均误差仅有0.0494%。用该模型预测2011-2020的人口总量,预测结果表明我国人口总数的增长已得到有效的控制。
Population is a complicated social issue, and an accurate population forecast is a focal point of social development plan of the nation, especially for formulating family planning policies. This paper on the one hand discusses the principle, structure and implementation method of BP neural network improved by the genetic algorithm, and on the other hand, by the use of the population data from 1982 to 2010 with a mean assay error of only 0.0494%, applies the improved model to predict the total population in 2011 to 2012, confirming the effectiveness of national control in population rise.
出处
《战略决策研究》
2012年第4期48-52,共5页
Journal of Strategy and Decision-Making
关键词
遗传算法
BP神经网络
人口预测
genetic algorithm
BP neural network
population forecast
