no code implementations • 23 Sep 2022 • P. Richard Hahn, Andrew Herren
What is the ideal regression (if any) for estimating average causal effects?
no code implementations • 15 Sep 2022 • Nikolay Krantsevich, Jingyu He, P. Richard Hahn
Determining subgroups that respond especially well (or poorly) to specific interventions (medical or policy) requires new supervised learning methods tailored specifically for causal inference.
1 code implementation • 21 Aug 2022 • Andrew Herren, P. Richard Hahn
We use this connection to show that challenges in SHAP approximations largely relate to the choice of a feature distribution and the number of $2^p$ ANOVA terms estimated.
no code implementations • 23 Apr 2022 • Meijiang Wang, Jingyu He, P. Richard Hahn
Despite this success, standard implementations of BART typically provide inaccurate prediction and overly narrow prediction intervals at points outside the range of the training data.
1 code implementation • 14 Sep 2020 • Andrew Herren, P. Richard Hahn
According to this formulation, large unlabeled data sets could be used to estimate a high dimensional propensity function and causal inference using a much smaller labeled data set could proceed via weighted estimators using the learned propensity scores.
1 code implementation • 9 Feb 2020 • Jingyu He, P. Richard Hahn
This paper develops a novel stochastic tree ensemble method for nonlinear regression, which we refer to as XBART, short for Accelerated Bayesian Additive Regression Trees.
no code implementations • 4 Oct 2018 • Jingyu He, Saar Yalov, P. Richard Hahn
Bayesian additive regression trees (BART) (Chipman et.
3 code implementations • 25 Sep 2018 • Ruocheng Guo, Lu Cheng, Jundong Li, P. Richard Hahn, Huan Liu
This work considers the question of how convenient access to copious data impacts our ability to learn causal effects and relations.
no code implementations • 14 Jun 2018 • P. Richard Hahn, Jingyu He, Hedibert Lopes
This paper develops a slice sampler for Bayesian linear regression models with arbitrary priors.
4 code implementations • 29 Jun 2017 • P. Richard Hahn, Jared S. Murray, Carlos Carvalho
This paper develops a semi-parametric Bayesian regression model for estimating heterogeneous treatment effects from observational data.
Methodology
no code implementations • 1 May 2014 • Tom LaGatta, P. Richard Hahn
We construct a stochastic version of the OLS estimator, which is a continuous disintegration exactly for the class of "uncorrelated implies independent" (UII) measures.