Causal Inference Methods In Python
In this project, I work in a team of four, we evaluate three causal inference algorithms to compute the average treatment effect (ATE) on two distinct datasets and compare their computational efficiency and performance. One dataset contains high dimensional data and another contains low dimensional data. We will use L1 penalized logistic regression to estimate the propensity scores for these two datasets then apply the three methods to calculate ATE for each dataset.
I developed and performed the Stratification algorithm, which had an accuracy of 86% on the low dimensional dataset, and 91% on the high dimensional dataset.
See the repo
This project was part of the class Applied Data Science (STAT GR5243) which I took in Spring 2021.
Explore more like this
The Two Year Long Process From Exploratory Data Analysis to Story
This while the majority of this analysis was conducted in August 2019, it this notebook was revisited and revised in December 2021 for use in a Eye Story. The resulting...
Plotting Racial Inequity In Higher Education with College Scorecard Data
This began with an exploratory data analysis on data provided by the Government College Scorecard. In this analysis I charted the popularity of different majors across the Ivy’s from 1998-2019,...
