A collection of my projects
For my Data Science II class, my group used LASSO, Elastic Net, Ridge Regression, Partial Least Squares, and MARS to predict COVID-19 recovery times and identify risk factors for long COVID.
For my causal inference capstone, I simulated data and used an incorrectly specified outcome regression model and a correctly specified inverse probability weighting model to show the robustness of doubly robust estimation (and vice-versa.)
For my mathematical statistics capstone, my group gave a presentation on sufficiency and the factorization theorem. After our presentation, we handed out the following worksheet to connect sufficiency to other properties of estimators and the Rao-Blackwell theorem.
For my survival analysis capstone, we used a log-normal and Weibull AFT model to find factors that affect survival rates of brain cancer patients. We used Schoenfeld residual plots to asses the PH assumption of our models, and then we compared our models using their AIC values.
For my correlated data capstone, my group used generalized estimating equations to predict ridership in Washington D.C.