A Bayesian nonparametric approach to causal inference on quantiles.
A Bayesian nonparametric approach to causal inference on quantiles.
复制标题
DOI：
10.1111/biom.12863
复制
发表时间：
2018-09
期刊：
影响因子：
1.9
通讯作者：
Winterstein AG
中科院分区：
文献类型：
--
作者：
Xu D;Daniels MJ;Winterstein AG
We propose a Bayesian nonparametric approach (BNP) for causal inference on quantiles in the presence of many confounders. In particular, we define relevant causal quantities and specify BNP models to avoid bias from restrictive parametric assumptions. We first use Bayesian additive regression trees (BART) to model the propensity score and then construct the distribution of potential outcomes given the propensity score using a Dirichlet process mixture (DPM) of normals model. We thoroughly evaluate the operating characteristics of our approach and compare it to Bayesian and frequentist competitors. We use our approach to answer an important clinical question involving acute kidney injury using electronic health records.
登录
查看更多内容
影响因子：
6.3
作者：
Frandsen, Brigham R.;Froelich, Markus;Melly, Blaise
通讯作者：
Melly, Blaise
DOI：
10.1080/01621459.2016.1155993
发表时间：
2017-06-01
影响因子：
3.7
作者：
Linn, Kristin A.;Laber, Eric B.;Stefanski, Leonard A.
通讯作者：
Stefanski, Leonard A.
影响因子：
3.7
作者：
Chung Y;Dunson DB
通讯作者：
Dunson DB
影响因子：
4.1
作者：
Gomes, Diane M.;Smotherman, Carmen;Jankowski, Christopher A.
通讯作者：
Jankowski, Christopher A.
影响因子：
6.3
作者：
Chib, Siddhartha
通讯作者：
Chib, Siddhartha