SimBaRepro
Simulation-based, Finite-sample Inference for Privatized Data
In 2024-25, I had the honor of working with Prof. Jordan Awan and Zhanyu Wang to turn their research paper into an R package published on CRAN.
This figure shows a rough description of the main idea behind the paper by Awan and Wang.
Through simulation-based methods, users can obtain p-values for hypothesis tests, confidence intervals, and multivariate confidence regions. Multivariate confidence regions can also be projected down to two-dimensional confidence region before they can be visualized.
The figure on the left shows a two-dimensional confidence region (1 indicates inclusion), and we can visualize it easily with the built-in `plot_grid` function (figure on the right).
To use the package or to see more details, see SimBaRepro.