Calendar
Course Overview
Privacy Attacks
Definitions and Basic Techniques
Definitions and Basic Techniques
- Sept 16
- Laplace Mechanism and Properties of DP
- Note, Slides
- Sept 18
- Properties of DP, Selection problem
- Note, Slides
- Sept 20
- Recitation
- Review on Laplace, DP properties, and HW1
Properties and Mechanisms
- Sept 23
- Exponential mechanism
- Slides, Note
- Sept 25
- Exponential mechanism and report noisy max
- Slides, Note
- Sept 27
- Recitation
- Review on Exp. Mech.
Other notions of DP and Gaussian Mechanism
- Sep 30
- Approximate DP, Gaussian Mechanism (Part 1)
- Slides, Note
- Oct 2
- Approximate DP, Gaussian Mechanism (Part 2)
- Slides, Note
- Oct 4
- Recitation
- Review on Approx. DP, Gauss. Mech., HW 2
- Oct 7
- Approximate DP and Advanced Composition
- Slides, Note
- Oct 9
- Sparse vector technique, Summary of mechanisms, Intro to ML
- Slides, Note
- Oct 11
- Recitation
- Review on Advanced Composition, HW 1&2
Differential Privacy and Machine Learning
Differential Privacy and Machine Learning
- Oct 28
- Wrap-Up Privacy Attacks on ML, Intro to Optimization
- Slides, Note
- Oct 30
- DP-SGD
- Slides, Note, HW3
- Nov 1
- Recitation
- Review on ML and DP-SGD
Other models of DP ML
- Nov 4
- Class Canceled
- Nov 6
- Review DP-SGD; Amplification by subsampling; PATE
- Slides
- Nov 8
- Recitation
- Review on DP-SGD, PATE and HW 3
- Nov 11
- Local DP
- Slides
- Nov 13
- Federated learning, Local DP + Multi-party computation
- Reading, Slides
- Nov 15
- Recitation
- Review on Midterm & HW3
- Nov 18
- Private synthetic data, Review of private ML
- Slides
Fairness
- Nov 20
- Fairness Metrics (Part 1)
- Slides
- Nov 25
- Fairness Metrics (Part 2)
Course Review and Final Exam
- Dec 2
- Course Review
- Dec 4
- Final Exam