Skip to content

mostly-ai/paper-fidelity-accuracy

Repository files navigation

Holdout-Based Empirical Assessment of Mixed-Type Synthetic Data

Reference implementation, plots and datasets accompanying the paper "Holdout-Based Empirical Assessment of Mixed-Type Synthetic Data" published at https://www.frontiersin.org/articles/10.3389/fdata.2021.679939/full.

Additional visualizations of the empirical study can be found at https://public.tableau.com/profile/michael.platzer1903#!/vizhome/paper-benchmarks/tradeoff.

2023-05 Update

Up-to-date benchmarks and evaluation were added to the 2023-05 sub-folder. You can launch the evaluation Jupyter Notebook directly on Google Colab here.

Acknowledgments

This research is supported by the "ICT of the Future” funding programme of the Austrian Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology. See https://anonymousbigdata.net/ for further details.

About

public repo for paper on fidelity / privacy assessment of mixed-type synthetic data

Resources

License

Stars

4 stars

Watchers

4 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors