Projects

Current

Applications of Amortized Bayesian Inference

Leveraging simulation-based inference with normalizing flows to enable tractable Bayesian estimation of differentiated product demand systems and microeconometric equilibrium models.

Measuring causal effects of lead in drinking water

with Ludovica Gazze, Wes Austin & Jarat Kochhar

Exploiting temporal variation in chloramination in public water systems to measure effects of corrosion-induced lead release in US drinking water on educational and health outcomes.


Past

Bayesian Double Machine Learning for Causal Inference

Research Assistance for Frank diTraglia & Laura Liu

Comparing a novel Bayesian Double Machine Learning approach in the partial linear model to frequentist double machine learning and Bayesian alternatives for causal inference in high-dimensional settings.

Simulation study

The Shapley Value in Machine Learning and the SHAP Value in Economics

On the interpretation of covariate attributions for causal forest predictions of heterogeneous treatment effects.

Mapping Disparities in Childhood Lead Exposure in England

with Frank diTraglia & Ludovica Gazze

Modeling prevalence of BLL elevation in England using Bayesian methods. Discontinued in favour of ECLIPS.