Mathematics
2 years ago

New Privacy-Preserving Method Safeguards Sensitive Data in Causal Inference Studies

1 view
arXiv (Cornell University)
Privacy-Preserving Causal Inference via Inverse Probability Weighting
Si Kai Lee, Luigi Gresele, Mijung Park, Krikamol Muandet

Paper Summary

Inverse probability weighting (IPW) methods are commonly used to estimate treatment effects in various fields. A new framework called privacy-preserving IPW (PP-IPW) has been developed to protect sensitive information in these studies. The PP-IPW framework was tested on different datasets and showed consistent results with the theoretical analysis.

Related papers

Theme background image
Mathematics
a year ago

New methods in R revolutionize causal inference with observational data.

Causal Inference
3 views
Theme background image
Mathematics
3 years ago

Inverse probability weighting may not correct bias in cohort studies' attrition.

Causal Inference
Social Factors
Missing Data
1 view
Theme background image
Mathematics
3 years ago

New study reveals pitfalls in using propensity scores for causal inference.

Causal Inference
Missing Data
Regularization Methods
2 views