Here are the experiments for the paper: Confidence Intervals for Private Query Processing. The experiments are not complicated so I simplily present them in jupyter notebooks.
- Input data are stored in bank_marketing.csv adult.csv and HepPh.txt.
-
Synthetic.ipynb describes results on synthetic datasets
-
medianCI.ipynb describes the results on real datasets
- meanCI.ipynb describes the experiments for mean CI.
- Ladder.ipynb describes experiments on graph counting queries
- Residual.ipynb describes the experiments on TPC-H queries