[ABE-L] Academic Seminar of Data Science with Jared Murray of University of Texas at Austin

[Hedibert Lopes]

Data Science -  Academic Seminar - INSPER


November 04, 2021, 12pm de São Paulo, Brasil (UTC/GMT -03:00)


https://zoom.us/j/95660781314


Title: Bayesian tree models with targeted smoothing for causal inference


Speaker: Jared Murray, University of Texas at Austin


 Abstract:

 Bayesian tree models like Bayesian additive regression trees (BART) and
Bayesian causal forests (BCF) are popular and effective methods for
inferring heterogeneous causal effects. However, their function estimates
are necessarily discontinuous and "rough" in their arguments, a significant
disadvantage in applications involving continuous  treatments or effect
moderators thought to have smoothly evolving relationships with treatment
efficacy. Here we  extend Bayesian tree models with "targeted smoothing" to
allow for (possibly) irregularly spaced continuous treatment

 variables or moderators while maintaining computational efficiency through
the use of carefully constructed basis expansions.


This talk will draw on the following papers:


 P. Richard Hahn. Jared S. Murray. Carlos M. Carvalho. "Bayesian Regression
Tree Models for Causal Inference: Regularization, Confounding, and
Heterogeneous Effects (with Discussion)." Bayesian Anal. 15 (3) 965 - 1056,
September 2020. https://doi.org/10.1214/19-BA1195


Jennifer E. Starling. Jared S. Murray. Carlos M. Carvalho. Radek K.
Bukowski. James G. Scott. "BART with targeted smoothing: An analysis of
patient-specific stillbirth risk." Ann. Appl. Stat. 14 (1) 28 - 50, March
2020. https://doi.org/10.1214/19-AOAS1268


Spencer Woody, CM Carvalho, PR Hahn, JS Murray "Estimating heterogeneous
effects of continuous exposures using Bayesian tree ensembles: revisiting
the impact of abortion rates on crime" https://arxiv.org/abs/2007.09845
-------------- Pr�xima Parte ----------
Um anexo em HTML foi limpo...
URL: <http://lists.ime.usp.br/pipermail/abe/attachments/20211103/af5ac74e/attachment.htm>