Completed
Differentially private GANS (DP-GAN)
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Generative Adversarial Models for Privacy and Fairness
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Zero-sum games
- 3 Generative Adversarial Networks (GANS)
- 4 Why not using GANs as they are?
- 5 Membership inference attacks for GANS
- 6 Differential Privacy (DP) to the rescue!
- 7 TensorFlow Privacy
- 8 Differentially private GANS (DP-GAN)
- 9 DP-GAN: noisy Wasserstein GAN
- 10 DP-GAN results
- 11 Context-aware fair data publishing
- 12 Empirical risk minimization with MI?
- 13 Generative Adversarial Privacy & Fairness (GAPF)
- 14 Example: GAPF under log-loss
- 15 Data-driven GAPF
- 16 Penalty method
- 17 Real-life data: GENKI dataset
- 18 Adversary's neural network
- 19 Feedforward Neural Network (FNN) encoder
- 20 Transposed Convolution Neural Network (TCNNP) encoder
- 21 GENKI fairness vs utility
- 22 Siamese-GAPF (S-GAPF) What if the sensitive label can take many values?
- 23 Real-life data: HAR dataset
- 24 HAR fairness vs utility
- 25 Gaussian mixture data model
