摘要
人口数量模型对自然环境和经济等方面的影响重大,为寻找合适的人口数量模型,分别使用LASSO方法和逐步回归方法进行变量选择,再使用简单线性回归模型分别对LASSO方法和逐步回归方法挑选出来的变量进行拟合分析,发现两个模型的拟合效果都较差;故使用部分线性可加模型对挑选出来的影响因子进行拟合分析,得到LASSO方法挑选出的变量进行回归拟合结果较理想,并得到最优人口数量模型.由部分线性可加模型结果发现,人口数量主要由普通高中毕业生数、就业人数和医疗卫生机构数表达.此外,普通高中毕业生数与人口数量呈线性正相关关系;就业人数和医疗卫生机构数与人口数量呈非线性关系;就业人数呈先加速上升后减速上升的情形,医疗机构数总体上随着人口数量增长,数量有所上升,但上升的速度缓慢,直至后面逐渐趋于饱和,所得结果基本与现实情况相符.
The population model has a great impact on all aspects of society. In order to find a suitable population model,Lasso method and stepwise regression method are used to select the variables. Then,the variables selected by the Lasso method and the stepwise regression method were fitted by the Simple linear regression model,and the results showed that the fitting effect of both models was poor. So we use the partial linear additive model to fit the selected factors,and we find that the result of the model selected by Lasso method is better. From the optimal model,we also find that the population is mainly expressed by the number of high school graduates,the number of employment and the number of health care institutions. In addition,there is a linear positive correlation between the number of high school graduates and the number of population;the number of employed persons and the number of medical and health institutions have a non-linear relationship with the number of population;In general,the number of medical institutions increased as the population increased,but the increase was slow until it reached saturation,and the results were basically in line with the reality.
作者
麦继芳
刘惠
赵海清
MAI Ji-fang;LIU Hui;ZHAO Hai-qing(School of Mathematics and Statistics,Lingnan Normal University,Zhanjiang 524048,Guangdong,China)
出处
《韶关学院学报》
2020年第3期5-10,共6页
Journal of Shaoguan University
基金
广东省科技攻关计划项目(2017A030303085).
