比例数据基于Tobit分位数回归模型的贝叶斯变量选择

doi:10.16381/j.cnki.issn1003-207x.2019.0291

中国管理科学 ›› 2022, Vol. 30 ›› Issue (4): 63-73.doi: 10.16381/j.cnki.issn1003-207x.2019.0291

比例数据基于Tobit分位数回归模型的贝叶斯变量选择

赵为华¹, 王玲¹, 程喆¹, 张日权²

1. 南通大学理学院，江苏南通226019； 2.华东师范大学统计学院,上海200241

收稿日期:2019-03-05 修回日期:2019-07-30 出版日期:2022-04-20 发布日期:2022-04-26
通讯作者: 赵为华(1978-), 男(汉族), 江苏海门人, 南通大学理学院，教授, 统计学博士,研究方向:复杂数据统计建模及其应用,Email:zhaowhstat@163.com. E-mail:zhaowhstat@163.com
基金资助:
国家社会科学基金资助项目(15BTJ027)；国家自然科学基金资助项目(11971171)

Variable Selection of Proportional Data Based on Tobit Quantile Regression Model

ZHAO Wei-hua¹, WANG Ling¹, CHENG Zhe¹, ZHANG Ri-quan²

1. SchoolofSciences, Nantong University, Jiangsu Nantong 226019, China;2. School of Statistics, East China Normal University, Shanghai 200241, China

Received:2019-03-05 Revised:2019-07-30 Online:2022-04-20 Published:2022-04-26
Contact: 赵为华 E-mail:zhaowhstat@163.com

摘要/Abstract

摘要： 针对家庭商业健康保险参保比例在［0,1］闭区间上取值的特点, 本文基于Tobit模型给出了比例响应数据的贝叶斯分位数回归建模方法。通过引入回归系数的“Spike-and-slab”先验分布, 应用EM 算法我们提出了基于门限规则的贝叶斯变量选择方法。大量数值模拟研究验证了所提的贝叶斯变量选择方法的有效性, 且具有易操作、计算量小等优点。最后, 将此方法应用到家庭商业健康保险数据的实证分析, 研究不同分位数水平下家庭健康保险参保比例的影响因素, 得到了许多有意义的研究结果。

关键词: 比例数据；Tobit分位数回归；贝叶斯变量选择; “Spike-and-Slab”先验; EM算法

Abstract: According to the features of the proportional data taken values on the closed interval ［0,1］ for the household coverage rate of commercial health insurance, in this paper, the Bayesian quantile regression modeling method is investigated for proportional response data based on the Tobit model. By introducing the “spike-and-slab” prior distributions of coefficients, the threshold rule-based Bayesian variable selection method is proposed by applying the EM algorithm. Extensive numerical simulation studies are used to verify the effectiveness of the proposed Bayesian variable selection method, and it has many merits such as easy implementation and cheap computation. Finally, the newly proposed method is applied to analyze the household coverage rate of commercial health insurance data, the influence factors of the coverage rate of commercial health insurance are uncovered at different quantile levels and some interesting results are gotten.

Key words: proportional data; tobit quantile regression; Bayesian variable selection; “Spike-and-slab” prior; EM algorithm

中图分类号:

O212

赵为华, 王玲, 程喆, 张日权. 比例数据基于Tobit分位数回归模型的贝叶斯变量选择[J]. 中国管理科学, 2022, 30(4): 63-73.

ZHAO Wei-hua, WANG Ling, CHENG Zhe, ZHANG Ri-quan. Variable Selection of Proportional Data Based on Tobit Quantile Regression Model[J]. Chinese Journal of Management Science, 2022, 30(4): 63-73.

参考文献

［1］ Ferrari S. Cribari-Neto F. Beta regression for modelling rates and proportions［J］. Journal of Applied Statistics, 2004, 31(7): 799-815.
［2］ Zhao Weihua, Zhang Riquan, Lv Yazhao. et al. Variable selection for varying dispersion beta regression model［J］. Journal of Applied Statistics, 2014, 41(1): 95-108.
［3］方匡南,姚紫薇. 半参数可加beta回归模型及其应用［J］. 中国管理科学, 2017, 25(9): 120-128.Fang Kuangnan, Yao Ziwei. Semi-parametric additive beta regression model with its application［J］.Chinese Journal of Management Science, 2017, 25(9): 120-128.
［4］ Song Xuekun, Tan Ming. Marginal models for longitudinal continuous proportional data［J］. Biometrics, 2000, 56(2): 496-502.
［5］ Smithson M. A better lemon squeezer? maximum-likelihood regression with beta-distributed dependent variables［J］. PsychologicalMethdos, 2006, 11: 54-71.
［6］ Rocha A, Simas A. Influence diagnostics in a general class of beta regression model［J］. Test, 2011, 20: 95-119.
［7］张朋, 邱振国. 基于单纯形分布的比例数据的回归分析［J］. 中国科学:数学, 2014, 44(1):89-104.Zhang Peng, Qiu Zhenguo. Regression analysis of proportional data using simplex distribution［J］. Sci Sin Math, 2014, 44(1):89-104.
［8］ Papke L, Woooldridge J. Econometric methods for fractional response variables with an application to 401(k)plan participation rates［J］. Journal of Applied Econometrics, 1996, 11(6): 619-632.
［9］ Zhao Weihua, Lian Heng, Zhang Riquan, et al. Estimation and variable selection for proportional response data with partially linear single-index models［J］. Computational Statistics & Data Analysis, 2016, 96: 40-56.
［10］ Ospina R, Ferrari S. A general class of zero-or-one inflated beta regression models［J］. Computational Statistics and Data Analysis, 2012, 56, 1609-1623.
［11］ Fang Kuangnan, Ma Shungge. Three-part model for fractional response variables with application to Chinese household health insurance coverage［J］. Journal of Applied Statistics,2013, 40(5): 925-940.
［12］ Bruno S, Heleno B. Bayesian analysis for zero-or-one inflated proportion data using quantile regression［J］.Journal of Statistical Computation and Simulation, 2015, 85(17): 1-15.
［13］ Koenker R, Bassett G. Regression quantiles［J］.Econometrica, 1978, 46(1): 33-50.
［14］ Koenker R. Quantile regression［M］. Cambridge,Cambridge University Press, 2005
［15］丁飞鹏. 固定效应部分线性单指数面板模型的惩罚分位数回归［J］. 统计研究, 2017, 34(2): 101-109.Ding Feipeng. Penalized quantile regression for partially linear single index panel models with fixed effect［J］. Statistical Research, 2017, 34(2): 101-109.
［16］张晶, 王淼晗, 方匡南. 我国城乡居民收入影响因素研究-基于省际面板分位数回归分析［J］. 数理统计与管理, 2015, 34(4): 571-579.Zhang Jing, Wang Miaohan, Fang Kuangnan. The determinants of income of resident in urban and rural areas-based on the provincial panle data quantile regression［J］. Journal of Applied Statistical and Management, 2015, 34(4): 571-579.
［17］李坤明,方丽婷. 空间滞后分位数回归模型的工具变量估计及参数检验［J］. 统计研究, 2018, 35(10):103-115.Li Kunming, Fang Liting. Instrumental variables estimation and parametric tests of spatial lag quantile regression model［J］. Statistical Research, 2018, 35(10): 103-115.
［18］陈耀辉, 郭俊峰, 殷文超. 人民币升值对中小板市场波动的影响——基于贝叶斯分位数回归的分析［J］. 系统工程, 2015(1): 39-47.Chen Yaohui, Guo Shaochao, Yin Wenchao. The impact of RMB appreciation on the volatility of small and medium-size board market based on bayesian quantile regression［J］. Systems Engineering, 2015(1): 39-47.
［19］赵为华, 张日权. 比例响应数据的分位数回归建模［J］. 数理统计与管理, 2017, 36(1): 73-84.Zhao Weihua, Zhang Riquan. Quantile regression modelling of proportional response data. Journal of Applied Statistical and Management, 2017, 36(1): 73-84.
［20］ Tobin J. Estimation of relationships for limited dependent variables［J］. Econometrica, 1958, 26(1): 24-36.
［21］ Mcdonald J, Moffitt R. The uses of tobit analysis［J］. Review of Economics and Statistics, 1980, 62(2): 318-321.
［22］ George E, Mcculloch R. Approaches for bayesian variable selection［J］. Statistica Sinica, 1997, 7(2): 339-373.
［23］ Yu K, Moyeed R. Bayesian quantile regression［J］. Statistics and Probability Letters, 2001, 54(4): 437-447.
［24］ Rockova V, George E. EMVS: The EM approach to bayesian variable selection［J］.Journal of the American Statistical Association, 2014, 109(506): 828-846.
［25］ Zhao Kaifeng, Lian Heng. Bayesian tobit quantile regression with single-index models［J］. Journal of Statistical Computation and Simulation, 2015, 85: 1247-1263.

[1]	袁先智, 周云鹏, 严诚幸, 刘海洋, 钱国骐, 王帆, 韦立坚, 李志勇, 李波, 李祥林, 曾途. 财务欺诈风险特征筛选框架的建立和应用[J]. 中国管理科学, 2022, 30(3): 43-54.
[2]	莫东序, 郑田丹. 基于复杂网络的投资组合优化研究[J]. 中国管理科学, 2021, 29(5): 25-33.
[3]	李庆, 张虎. 单指标非参数期权定价——改进的非参数定价方法[J]. 中国管理科学, 2020, 28(10): 43-53.
[4]	苑慧玲, 徐路, 周勇. 带有市场交易信息和随机微观噪声下的杠杆效应研究[J]. 中国管理科学, 2020, 28(9): 12-22.
[5]	陈强, 龚玉婷. 基于近邻截断的跳跃扩散过程波动函数的设定检验[J]. 中国管理科学, 2020, 28(7): 45-56.
[6]	许保光, 刘倩倩, 高敏刚. 基于机场繁忙程度的航班延误波及分析[J]. 中国管理科学, 2019, 27(8): 87-95.
[7]	何迪, 周勇. 基于状态空间模型的宏观经济因素对股市流动性的建模分析[J]. 中国管理科学, 2019, 27(5): 42-49.
[8]	蒋青嬗, 黄灿, 李毅君. 半参数空间ZISF的估计、分类与蒙特卡罗模拟[J]. 中国管理科学, 2019, 27(3): 20-29.
[9]	傅强, 张兴敏. 时变参数模型的最优滚动窗宽选择标准及应用[J]. 中国管理科学, 2018, 26(8): 20-30.
[10]	王江涛, 周勇. 高频数据瞬时波动率核估计的窗宽选择及算法研究[J]. 中国管理科学, 2018, 26(7): 1-8.
[11]	张莉, 邓礼英, 周勇. 半参数alpha策略的反转效应研究[J]. 中国管理科学, 2016, 24(12): 30-38.
[12]	李战江, 迟国泰, 党均章. 基于copula的追随者银行的企业项目总体风险评价模型[J]. 中国管理科学, 2015, 23(1): 99-110.
[13]	黄金波, 李仲飞, 周先波. VaR与CVaR的敏感性凸性及其核估计[J]. 中国管理科学, 2014, 22(8): 1-9.
[14]	蒋春福, 杨宇宽. 混合椭球分布下证券组合的尾部条件方差[J]. 中国管理科学, 2013, 21(4): 17-26.
[15]	蒋家东, 冯允成. 基于协方差阵特征结构的新型控制图研究[J]. 中国管理科学, 2011, 19(3): 123-133.

比例数据基于Tobit分位数回归模型的贝叶斯变量选择

Variable Selection of Proportional Data Based on Tobit Quantile Regression Model

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 10