Gaussian-Differential-Privacy-on-Riemannian-Manifolds

Code repo for paper Gaussian Differential Privacy on Riemannian Manifolds (Neurips 2023) https://arxiv.org/abs/2311.10101

Simulation on the Euclidean space R

Simulation on the unit circle S^1

For simulations on the unit circle, run s1_simulation.R. It stores the simulation results in files begins with s1_result_.

To generate plot like Figure 1 the paper, run s1_plot.R. The plot is stored in s1.png.

Simulation on the unit sphere S^2

For simulations on the unit circle, run sphere_simulation.R. It stores the simulation results in files begins with sphere_compare_result_.

To generate plot like Figure 1 the paper, run sphere_plot.R. The plot is stored in sphere_compares.png.

To be continue...