基于LassoCV & R型聚类的东莞城市人口变化影响因素分析

人口老龄化问题已经成为我国当前经济、社会发展中面临的一项重大问题，在各大城市人才争夺战呈现白热化的趋势下，分析、发现城市人口吸引力影响因素具有一定现实意义。使用LassoCV从城市影响因素的高维度数据集中提取到合理、简单的影响因素变量集，同时保证模型的数值性能和泛化能力；使用R型聚类对与目标变量正向相关的影响因素进行分类，得到更为清晰的筛选变量集结构。最后基于研究发现提出提升东莞城市人口吸引力的若干建议。

Analysis of Factors Influencing Urban Population Change in Dongguan Based on LassoCV&R Cluster Analysis

The issue of aging population has become a major issue in China''s current economic and social development. With the fierce competition for talent in major cities, analyzing and discovering the factors affecting the attractiveness of urban population has certain practical significance. A reasonable and simple set of influencing factor variables from high-dimensional datasets of urban influencing factors is extracted using LassoCV, while the numerical performance and generalization ability of the model is ensured. The R-type clustering was used to classify the influencing factors positively correlated with the target variable, resulting in a clearer structure of the screening variable set.Finally, based on the findings, several suggestions are proposed to enhance the attractiveness of Dongguan''s population.