Instructors: Salil Vadhan, James Honaker, and Wanrong Zhang
Teaching Fellows: Daniel Alabi, Jayshree Sarathy, Michael Shoemate, and Connor Wagaman
Regular meeting time: Tuesdays & Thursdays 11:15-12:30, starting 1/25
Location: Harvard Science & Engineering Complex SEC LL2.221, 150 Western Ave, Allston, MA
Date | Title | Slides/Notes | Advance Reading |
---|---|---|---|
Tue 1/18 | Course Preview | pdf, Recording | |
Attacks on Privacy | |||
Tue 1/25 | Course Overview & Attacks on Privacy | ||
Thu 1/27 | Reconstruction Attacks | Tanner, Barbaro-Zeller, Narayanan-Shmatikov | |
Problem Set 1 (due Weds 2/2) | pdf, tex | ||
Section 1 | |||
Tue 2/01 | Reconstruction & Membership Attacks | Smith-Ullman Reconstruction lecture notes (Sec 1-2.1, Sec 3), Ruggles-van Riper, Jessica Hullman blog post | |
Thu 2/03 | Membership Attacks | pdf, ipynb | P3G Consortium et al., Korolova (§1,4,6,8) |
Problem Set 2 (due Fri 2/11) | pdf, tex | ||
Section 2 | |||
Foundations of Differential Privacy | |||
Tue 2/08 | Definition, basic mechanisms | DP Primer Secs III-IV.B | |
Thu 2/10 | DP Foundations: the Laplace Mechanism | pdf, ipynb | U.S. Broadband Coverage Data Set, Smith-Ullman Lecture 4 (Sec 4) |
Problem Set 3 (due Fri 2/18) | pdf, tex | ||
Section 3 | |||
Tue 2/15 | More DP Foundations | DP Primer Secs IV.C-VI.B | |
Thu 2/17 | The Gaussian Mechanism | pdf, ipynb wk4_* | Smith-Ullman Lecture 5 (Sec 1), Lecture 9 (Lemma 1.2, Sec 2 up to Thm 2.1), Lecture 10 (Sec 1) |
Problem Set 4 (due Fri 2/25) | pdf, tex | ||
Section 4 | |||
Tue 2/22 | Beyond Noise Addition | pdf, ipynb | Smith-Ullman Lecture 6 (Secs 1.0-1.1), Smith-Raskhodnikova encyclopedia article: DP for graph data |
Implementing Centralized DP | |||
Thu 2/24 | Synthetic Data: the 2020 Census | 2020 Census Data Products and Privacy Methods (Secs 2.0-2.2, 3, 5.0) | |
Problem Set 5 (due Fri 3/4) | pdf, tex | ||
Section 5 | |||
Tue 3/1 | Statistical Releases: the Opportunity Atlas | Opportunity Atlas; Chetty-Friedman JPC (Sec 3) | |
Tue 3/1 and Thu 3/3 | Machine Learning with DP | pdf, ipynb | Deep Learning with DP (Secs 1-3.1, 5.2) |
Problem Set 6 (due Fri 3/11) | pdf, tex | ||
Section 6 | |||
Tue 3/8 | Machine Learning with DP (cont’d) | pdf, ipynb | |
Thu 3/10 | Embedded EthiCS module | Nissenbuam, Privacy in Context (Intro and Ch 7) | |
Problem Set 7 (due Fri 3/25) | pdf, tex | ||
Embedded EthiCS Assignment (due Fri 3/25) | |||
Section 7 | |||
Tue 3/22 | Programming Frameworks and Query Systems | DP for DB (Secs 7.0, 7.2, 7.3, 7.6, 8.0, 8.1) | |
Thu 3/24 | OpenDP | pdf, ipynb | PSI (Secs 0-3) |
Problem Set 8a (due Fri 4/1) | pdf, tex | ||
Section 8 | ipynb | ||
Local and Distributed Models | |||
Tue 3/29 | Local Model: Foundations | ||
Thu 3/31 | Local Model: Practicum | Learning with Privacy at Scale | |
Section 9 | |||
Tue 4/05 | Local Model: Implementations | Federated Learning and Privacy | |
Thurs 4/07 | Other Distributed Models: Foundations | DP for DB (Chapter 9) | |
Problem Set 8b (due Fri 4/15) | |||
Section 10 | |||
Tues 4/12 | Other Distributed Models: Implementations | Privacy-Preserving RCT | |
Societal Perspectives on Privacy | |||
Thurs 4/14 | Perspectives on Privacy: Law and Policy | Nissim-Wood (Secs 2.1, 2.3), Solove (pgs. 479-83, 488-91), Cohen (pgs. 1915-20, 1926-27) | |
Tues 4/19 | Perspectives on Privacy: Science & Technology Studies | Winner (pgs. 121-124, 130-131) Green-Viljoen (Secs 1-2) Mulligan-Koopman-Doty (pgs. 1-5, 9-10), Sarathy (optional) | |
Thurs 4/21 | Industry and Government Panel | ||
Tues 4/26 | Conclusions | ||
Mon-Tues 5/9-10 | Project Presentations |