Evaluation of Differeentially private 2D histogram publishing algorithms

This project is to evaluate the 2D histogram publishing algorithms

Project structure:

• Algorithm: all algorithms
• Result: result of evalutation
• data_generator.py - to generate synthetic dataset
• evaluation.py - contains evaluation metrics
• experiment.py - pipeline to run experiment
• make_plot.py - functions to generate plot
• main.py - main program to run whole experiment

Algorithms

• Identity
• Uniform
• Uniform Grids
• Adaptive Grids
• Quad tree (uniform, geometric, leaf)
• K-D tree (uniform, geometric, leaf, cell-based)
• Homogenous Tree (HTF)