Differential privacy
A mathematical guarantee that the result of an analysis, or a model trained on data, changes very little whether or not any one person’s data is included, bounded by a privacy-loss parameter epsilon; a smaller epsilon means stronger privacy and noisier results. NIST SP 800-226 explains how to evaluate such claims.
Covered in
Guide
Tokenization, pseudonymization and privacy transformations: choose deliberately
Choose between vault tokens, keyed pseudonyms, format-preserving encryption, aggregation and differential privacy by the joins, reversibility and leakage each allows, and test re-identification on the data you release.
Guide
Assess extraction, membership and inference risks
Define what an adversary can observe and measure privacy leakage without confusing a model’s output with proof about its training data.