https://us02web.zoom.us/j/88018185733?pwd=OEcxb2ZSYUwzV2hWTE9wcE1yTWRUQT09
Passcode: 279596
Speaker: Professor Ioanna Manolopoulou, University College London
Title: "Bayesian Causal Forests for Heterogeneous Treatment Effects Estimation"
Abstract: Bayesian Causal Forests (BCF) are a recently proposed nonparametric causal regression model that employs Bayesian Additive Regression Trees and is specifically designed to estimate heterogeneous treatment effects using observational data (Hahn et al, 2020). In this talk we will discuss developments in BCF in two directions. In the first, we introduce adapting shrinkage across different covariates. The extended version presented in this work, which we name Shrinkage Bayesian Causal Forest, is equipped with an additional pair of priors allowing the model to adjust the weight of each covariate through the corresponding number of splits in the tree ensemble. These priors improve the model's adaptability to sparse data generating processes and allow to perform fully Bayesian featu.re shrinkage in a framework for treatment effects estimation, and thus to uncover the moderating factors driving heterogeneity. The second direction is to expand the model so that it can handle the combination of randomised and observational data, by allowing the two pieces of data to share some parts of the model. We show that careful consideration of the common and different characteristics of the two datasets can allow us to fine-tune the model and draw inferences that are more generalisable. Joint work with Alberto Caron, Ilina Yozova Gianluca Baio