Releasing DP synthetic data built from treatment-arm and outcome moments, plus noise-aware multiple imputation, gives calibrated ATE intervals at strict privacy budgets.
Model agnostic differentially private causal inference.arXiv preprint arXiv:2505.19589,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Workload-Preserving Differentially Private Synthetic Data for Causal Inference via Maximum-Entropy Calibration
Releasing DP synthetic data built from treatment-arm and outcome moments, plus noise-aware multiple imputation, gives calibrated ATE intervals at strict privacy budgets.