# Mitigating Statistical Bias within Differentially Private Synthetic Data

arXiv: https://arxiv.org/abs/2108.10934 · HTML: https://mlsanity.com/paper/2108.10934/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 55 references that are in the map.
- First posted on arXiv 24 Aug 2021.

## Builds on (most cited first, 24 of 24)

- [Deep Learning with Differential Privacy](https://mlsanity.com/paper/1607.00133/) (2016)
- [Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs](https://mlsanity.com/paper/1706.02633/) (2017)
- [Differentially Private Generative Adversarial Network](https://mlsanity.com/paper/1802.06739/) (2018)
- [Improving the Gaussian Mechanism for Differential Privacy: Analytical Calibration and Optimal Denoising](https://mlsanity.com/paper/1805.06530/) (2018)
- [A General Framework for Updating Belief Distributions](https://mlsanity.com/paper/1306.6430/) (2013)
- [DP-CGAN: Differentially Private Synthetic Data and Label Generation](https://mlsanity.com/paper/2001.09700/) (2020)
- [Private Selection from Private Candidates](https://mlsanity.com/paper/1811.07971/) (2018)
- [Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting](https://mlsanity.com/paper/1906.09531/) (2019)
- [Privacy for Free: Posterior Sampling and Stochastic Gradient Monte Carlo](https://mlsanity.com/paper/1502.07645/) (2015)
- [Discriminator Rejection Sampling](https://mlsanity.com/paper/1810.06758/) (2018)
- [Pareto Smoothed Importance Sampling](https://mlsanity.com/paper/1507.02646/) (2015)
- [Metropolis-Hastings Generative Adversarial Networks](https://mlsanity.com/paper/1811.11357/) (2018)
- [Differentially Private Mixture of Generative Neural Networks](https://mlsanity.com/paper/1709.04514/) (2017)
- [Leveraging Public Data for Practical Private Query Release](https://mlsanity.com/paper/2102.08598/) (2021)
- [Differentially Private Synthetic Data: Applied Evaluations and Enhancements](https://mlsanity.com/paper/2011.05537/) (2020)
- [On the Theory and Practice of Privacy-Preserving Bayesian Data Analysis](https://mlsanity.com/paper/1603.07294/) (2016)
- [General Bayesian Updating and the Loss-Likelihood Bootstrap](https://mlsanity.com/paper/1709.07616/) (2017)
- [Differentially Private Data Generative Models](https://mlsanity.com/paper/1812.02274/) (2018)
- [Differentially Private Generative Adversarial Networks for Time Series, Continuous, and Discrete Open Data](https://mlsanity.com/paper/1901.02477/) (2019)
- [Private Post-GAN Boosting](https://mlsanity.com/paper/2007.11934/) (2020)
- [Bayesian Differential Privacy through Posterior Sampling](https://mlsanity.com/paper/1306.1066/) (2013)
- [Privacy-preserving parametric inference: a case for robust statistics](https://mlsanity.com/paper/1911.10167/) (2019)
- [Differentially Private Releasing via Deep Generative Model (Technical Report)](https://mlsanity.com/paper/1801.01594/) (2018)
- [Foundations of Bayesian Learning from Synthetic Data](https://mlsanity.com/paper/2011.08299/) (2020)

## Built on by (most cited first, 4 of 4)

- [The Real Deal Behind the Artificial Appeal: Inferential Utility of Tabular Synthetic Data](https://mlsanity.com/paper/2312.07837/) (2023)
- [Debiasing Synthetic Data Generated by Deep Generative Models](https://mlsanity.com/paper/2411.04216/) (2024)
- [Differentially Private Statistical Inference through β-Divergence One Posterior Sampling](https://mlsanity.com/paper/2307.05194/) (2023)
- [A density ratio framework for evaluating the utility of synthetic data](https://mlsanity.com/paper/2408.13167/) (2024)

## Abstract (arXiv)

Increasing interest in privacy-preserving machine learning has led to new and evolved approaches for generating private synthetic data from undisclosed real data. However, mechanisms of privacy preservation can significantly reduce the utility of synthetic data, which in turn impacts downstream tasks such as learning predictive models or inference. We propose several re-weighting strategies using privatised likelihood ratios that not only mitigate statistical bias of downstream estimators but also have general applicability to differentially private generative models. Through large-scale empirical evaluation, we show that private importance weighting provides simple and effective privacy-compliant augmentation for general applications of synthetic data.