摘要
将半参数回归与广义可加模型结合,应用于居民个人收入影响因素研究。结果表明,在影响个人收入的诸因素中,个人能力、受教育年限、潜在经验和是否居住在城市对居民个人收入的影响是参数效应,影响均表现为正效应;而父母受教育的年限对居民个人收入影响是非参数效应。对比参数模型和半参数广义可加模型,半参数广义可加模型的实证结果解释能力更强,更加符合实际。
In this paper,semi-parametric regression and generalized additive models are combined to study the influencing factors of residents′personal income.The results show that factors such as ability,education,potential experience and whether living in the urban areas have positive parametric effects on the influence of personal income and the years of education for the parents is a non-parameter effect.By comparing parameter models and semi-parametric generalized additive models,the empirical results show the latter is more competent and practical.
作者
刁海涛
DIAO Haitao(School of Applied Mathematics,Nanjing University of Finance&Economics,Nanjing 210023,China)
出处
《宿州学院学报》
2019年第10期68-70,84,共4页
Journal of Suzhou University
基金
国家社会科学基金项目(16BTJ030)
江苏省研究生科研创新计划项目(KYCX17_1206)
关键词
半参数
广义可加模型
居民个人收入
Semi-parameter
Generalized additive model
Residents′personal income