Privacy-Preserving ML
Differential privacy accounting, federated learning, secure aggregation, and the utility cost of each guarantee.
5concepts
62flashcards
36minutes of reading
- 01 Differential Privacy and the Epsilon You Actually Get What the differential privacy guarantee says precisely, how DP-SGD achieves it through clipping and noise, and why the epsilon values used in practice permit far more leakage than the formalism's reputation suggests.
- 02 Federated Learning and Secure Aggregation Why keeping data on device is not by itself a privacy guarantee, how gradient inversion recovers training inputs, and what secure aggregation does and does not prevent.
- 03 Machine Unlearning What it would mean to remove a training example's influence from a model, why retraining is the only exact method, and the approximate approaches and their verification problem.
- 04 Synthetic Data and What It Does Not Protect Why generating data from a model trained on real records does not by itself anonymise anything, the utility-privacy frontier that no generator escapes, and the evaluation that a synthetic dataset needs.