In this theoretical work we provide an unrestricted linear model for identifying treatment effects with time-varying qualification with panel data and repeated cross-sections. This model enables fixed and random effects estimation of treatment effects; the latter has not been considered before. In addition we identify new scenarios and respectiv…

R code (Rmd and Rnw) of the analysis on treatment effect of income on religiousness (Busser 2015)

Experimental React component for an interactive spinning text treatment

Gaines and Kuklinski (2011) Estimators for Hybrid Experiments

A biologically motivated mathematical formalism is used to estimate the relative risks of breast, lung and thyroid cancers in childhood cancer survivors due to concurrent therapy regimen. This model specifically includes possible organ-specific interaction between radiotherapy and chemotherapy. The model predicts relative risks for developing se…

Comparison of treatment effect in Randomized Control Trial (RCT) and Propensity Score Matching methods, conducted on Large-Scale Dataset by 'Criteo'.

Lightweight uplift modeling framework for Python

Propensity Score based Matching via Distribution Learning

Treatment Effect Heterogeneity visualization using R

Find sub groups (segments) with heterogeneous treatment effect in Randomised Controlled Trial data.

This is a demonstration of how we can implement Hirano-Imbens (2004) model for estimating Average Dose Response Function under Normally distributed continuous treatment.

Quantifying the Persistence of Misinformation: A Case Study in R

Endogenous switching regression model

Review: Data-driven methodology for detecting treatment effect heterogeneity

Multiple Responses Subgroup Identification

This repository provides R-code for the estimation of the conditional average treatment effect (CATE) using machine learning (ML) methods.

Github for the final project in Econometrics. Replication of Cameron, Gelbach and Miller (2008) and extension to "skewed" treatment variable.

Implementation of neural network algorithm for estimation of heterogeneous treatment effects and propensity scores described in Farrell, Liang, and Misra (2021)

Code for the WSDM '20 paper, Learning Individual Causal Effects from Networked Observational Data.