Inspired by The Markup's investigative reporting into the Home Mortgage Disclosure Act, which reports all individual mortgage applications including characteristics like applicants' self-reported race, income, and loan amount, as well as the final approval/rejection decision, I wanted to explore the 2018 to 2020 data sets and understand how the pandemic may have affected approval rates.
Includes various exploratory data analysis exercises to understand trends in the 2020 HMDA data set.
Comprehensive culminating project on the 2018 to 2020 HMDA data sets, including descriptive statistics, data cleaning rationale, statistical tests, and findings. Includes final report and the associated Python code.
Supervised learning project in R to predict mortgage approval for all 2020 mortgage applications. Details data wrangling, model evaluation, and results.