Professionally, I have worked for almost 10 years in the field of data science for progressive political campaigns and labor unions in the United States. My main area of expertise is building microtargeting models where I assign a score of 0 to 100 to up to 25 million voters at a time predicted their likelihood to support an endorsed candidate, likelihood to vote in an election or likelihood to take an action. I develop these models mainly using logistic regression, random forest, lasso regression, and CART using Stata, R and Python. I also have developed forecasting techniques using Bayesian methodology and data wrangled very messy data in ...
Member since May 2020