作　　者： ;

机构地区： 广东工业大学自动化学院

出　　处： 《东莞理工学院学报》 2019年第5期27-32,44,共7页

摘　　要： 人口普查是政府在各个时期获取人口资料、掌握国情国力的一种最基本的调查方法。即使在现在科技发达的社会,人口普查也是非常耗时耗力的。自中华人民共和国成立,只进行过6次全国性的人口普查。在这个数据爆炸式增长、科技快速进步的时代,通过人工智能技术并借助大数据来估算城市人口,可以节省大量的时间成本和人力消耗,能够让人口普查工作更为高效。甚至有实现对人口的实时动态预测。运用LightGBM算法和ARIMA算法分别对人口流动情况进行预测,比较两个算法在人口流动预测应用方向的准确率,得出了LightGBM算法在人口流动预测上的准确率比ARIMA算法的准确率高,同时ARIMA算法的计算时间是LightGBM算法计算时间的7至10倍,甚至更多倍数。 The census is the most basic survey method for the government to obtain population data and master the national strength of the country at various times. The census is very time-consuming and labor-intensive. After the founding of New China, China has only conducted six national censuses. In this era of explosive data growth and rapid advancement of data technology, the use of artificial intelligence to estimate urban population through big data can enable census work to be completed more efficiently.This will save a lot of time and manpower, and it is even possible to achieve real-time dynamic population projections. In this paper, LightGBM algorithm and ARIMA algorithm are used to predict the population flow respectively, and the accuracy of the two algorithms in the application direction of population flow prediction is compared. It is concluded that the accuracy of LightGBM algorithm in population flow prediction is better than that of ARIMA algorithm, and the calculation time of ARIMA algorithm is 7 to 10 times or even more than that of LightGBM algorithm.

关 键 词： 算法 算法 人口流动预测

领　　域： []