Differentially Private Data Generation with Missing Data. (arXiv:2310.11548v1 [cs.DB])

Despite several works that succeed in generating synthetic data with
differential privacy (DP) guarantees, they are inadequate for generating
high-quality synthetic data when the input data has missing values. In this
work, we formalize the problems of DP synthetic data with missing values and
propose three effective adaptive strategies that significantly improve the
utility of the synthetic data on four real-world datasets with different types
and levels of missing data and privacy requirements. We also identify the
relationship between privacy …

data differential privacy high input missing privacy private private data problems quality strategies synthetic synthetic data work