yuxiangw / autodp

autodp: A flexible and easy-to-use package for differential privacy
Apache License 2.0
265 stars 53 forks source link

a few good tutorials for autodp use cases #19

Closed yuxiangw closed 3 years ago

yuxiangw commented 3 years ago
  1. An end-to-end example for linear regression with SSP and AdaSSP (including actual implementation of the algorithm and benchmarking on an actual dataset).

  2. A tutorial on how to represent differentially private deep learning algorithms in autodp. It covers four popular algorithms: NoisyGD, NoisySGD, PATE and PrivateKNN.

  3. A separate tutorial for representing PATE with autodp together with how to use the numerical calibrator.