摘要
模型平均以其稳健性好,
预测精度高等诸多优点获得了当代统计学和计量经 济学界的高度关注,
在经济、金融、生物、医学等领域有着广泛的 应用前景.
模型平均的发展方向主要包括贝叶斯模型平均(BMA)和频 率模型平均(FMA).
文章介绍了贝叶斯模型平均方法, 改进贝叶斯 模型平均权重的-概率方法,
以及频率模型平均方法, 并对BMA和FMA进行了理论上的比较,
然后通过仿真研究比较了上述模型平均方法在线性和广义线性模型下的有限样本性能.
Abstract
Model averaging has attracted much attention
in the field of statistics and econometrics due to many advantages
such as good robustness and high prediction accuracy, and has great
potential applications in many fields such as economics, finance,
biology and medicine. Model averaging develops mainly in two directions:
Bayesian model averaging (BMA) and frequentist model averaging
(FMA). In this paper, BMA, -probability method which improves the
weight of BMA and FMA are briefly introduced, and the theoretical
comparison between BMA and FMA is carried out, then the simulation
studies are used to compare the finite sample performance of the
model averaging methods mentioned above under linear and generalized
linear models.
关键词
频率模型平均,贝叶斯模型平均,权重, -概率,广义线性模型.
{{custom_keyword}} /
乔鸽, 周建红, 李新民.
广义线性模型下模型平均的比较研究. 系统科学与数学, 2021, 41(4): 1164-1180. https://doi.org/10.12341/jssms19399
QIAO Ge, ZHOU Jianhong, LI Xinmin.
A Comparative Study of Model Averaging for Generalized Linear Models. Journal of Systems Science and Mathematical Sciences, 2021, 41(4): 1164-1180 https://doi.org/10.12341/jssms19399
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}