From fc344fadbbc09257f1e726ffc83383e95898c1e5 Mon Sep 17 00:00:00 2001 From: Justin Hsu Date: Thu, 22 Aug 2019 18:39:38 -0500 Subject: [PATCH] Adjusting paper presentation format. Plan: fix topic for each lecture and two suggested papers each. --- website/docs/org.md | 46 ++++++++++++++---------------- website/docs/resources/readings.md | 22 ++++++++++++++ website/docs/schedule/lectures.md | 42 +++++++++++++-------------- 3 files changed, 65 insertions(+), 45 deletions(-) diff --git a/website/docs/org.md b/website/docs/org.md index e5086a4..7a4bcc8 100644 --- a/website/docs/org.md +++ b/website/docs/org.md @@ -25,49 +25,47 @@ These three components are detailed below. ### Paper presentations -**Paper discussions** are one of the main components of this course. Before -every presentation, you are expected to read the paper closely and understand -its significance, including (a) the main problem addressed by the paper, (b) the -primary contributions of the paper, and (c) how the authors solve the problem in -some technical detail. Of course, you are also expected to attend discussions -and actively participate in the discussion. +**Paper discussions** are one of the main components of this course. In groups +of two (or very rarely three), you will present 2-3 papers on a related topic +and lead the discussion; we will have presentations most Wednesdays and Fridays. +Your presentation should last about **60 minutes** long, leaving the remainder +of the time for a wrap-up discussion. Please sign up for a slot and a paper by +**Monday, September 9**; while we will try to accommodate everyone's interests, +we may need to adjust the selections for better balance and coverage. -The topics we will be reading about are from the recent research -literature---peer-reviewed and published, but not completely refined. Most +Before every presentation, all students are expected to read the papers closely +and understand their significance, including (a) the main problems, (b) the +primary contributions, and (c) how the technical solution. Of course, you are +also expected to attend discussions and actively participate in the discussion. + +We will be reading about topics from the recent research literature. Most research papers focus on a very narrow topic and are written for a very specific technical audience. It also doesn't help that researchers are generally not the clearest writers, though there are certainly exceptions. These [notes](https://web.stanford.edu/class/ee384m/Handouts/HowtoReadPaper.pdf) by Srinivasan Keshav may help you get more out of reading papers. -To help you prepare for the class discussions, I will also send out a few -questions at least 24 hours before every paper presentation. **Before** each -lecture, you should send me brief answers---a short email is fine, no more than -a few sentences per question. These questions will help you check that you have -understood the papers---they are not meant to be very difficult or -time-consuming and they will not be graded in detail. - ### Homeworks There will be three small homework assignments, one for each of the core -modules. You will play with software implementations of the methods we cover in -class. These assignments are not weighted heavily, though they will be lightly -graded; the goal is to give you a chance to write some code. +modules, where you will play with software implementations of the methods we +cover in class. These assignments will be lightly graded; the goal is to give +you a chance to write some code and run some experiments. ### Course Project -The main component is the **course project**. You will work individually or in -pairs on a topic of your choice, producing a conference-style write-up and -presenting the project at the end of the semester. Successful projects may have -the potential to turn into an eventual research paper or survey. Details can be -found [here](assignments/project.md). +The main course component is the **course project**. You will work individually +or in pairs on a topic of your choice, producing a conference-style write-up and +presenting the project at the end of the semester. The best projects may +eventually lead to a research paper or survey. Details can be found +[here](assignments/project.md). ## Learning Outcomes By the end of this course, you should be able to... - Summarize the basic concepts in differential privacy, applied cryptography, - language-based security, and adversarial machine learning. + and adversarial machine learning. - Use techniques from differential privacy to design privacy-preserving data analyses. - Grasp the high-level concepts from research literature on the main course diff --git a/website/docs/resources/readings.md b/website/docs/resources/readings.md index 9f60daa..7824d93 100644 --- a/website/docs/resources/readings.md +++ b/website/docs/resources/readings.md @@ -24,6 +24,13 @@ ASIACRYPT 2015. - Matthew Joseph, Aaron Roth, Jonathan Ullman, and Bo Waggoner. [*Local Differential Privacy for Evolving Data*](https://arxiv.org/abs/1802.07128). + NIPS 2018. +- Albert Cheu, Adam Smith, Jonathan Ullman, David Zeber, and Maxim Zhilyaev. + [*Distributed Differential Privacy via Shuffling*](https://arxiv.org/pdf/1808.01394). + EUROCRYPT 2019. +- Jingcheng Liu and Kunal Talwar. + [*Private Selection from Private Candidates*](https://arxiv.org/pdf/1811.07971). + STOC 2019. ### Adversarial Machine Learning - Christian Szegedy, Wojciech Zaremba, Ilya Sutskever, Joan Bruna, Dumitru Erhan, Ian Goodfellow, and Rob Fergus. @@ -47,6 +54,12 @@ - Aleksander Madry, Aleksandar Makelov, Ludwig Schmidt, Dimitris Tsipras, and Adrian Vladu. [*Towards Deep Learning Models Resistant to Adversarial Attacks*](https://arxiv.org/pdf/1706.06083.pdf). ICLR 2018. +- Vitaly Feldman. + [*Does Learning Require Memorization? A Short Tale about a Long Tail*](https://arxiv.org/pdf/1906.05271). + arXiv 2019. +- Nicholas Carlini, Chang Liu, Ăšlfar Erlingsson, Jernej Kos, and Dawn Song. + [*The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks*](https://arxiv.org/pdf/1802.08232). + USENIX Security 2019. ### Applied Cryptography - Benjamin Braun, Ariel J. Feldman, Zuocheng Ren, Srinath Setty, Andrew J. Blumberg, and Michael Walfish. @@ -73,6 +86,15 @@ - Arjun Narayan, Ariel Feldman, Antonis Papadimitriou, and Andreas Haeberlen. [*Verifiable Differential Privacy*](https://www.cis.upenn.edu/~ahae/papers/verdp-eurosys2015.pdf). EUROSYS 2015. +- Henry Corrigan-Gibbs and Dan Boneh. + [*Prio: Private, Robust, and Scalable Computation of Aggregate Statistics*](https://people.csail.mit.edu/henrycg/files/academic/papers/nsdi17prio.pdf). + NSDI 2017. +- Valerie Chen, Valerio Pastro, Mariana Raykova. + [*Secure Computation for Machine Learning With SPDZ*](https://arxiv.org/pdf/1901.00329). + NIPS 2018. +- Wenting Zheng, Raluca Ada Popa, Joseph E. Gonzalez, Ion Stoica. + [*Helen: Maliciously Secure Coopetitive Learning for Linear Models*](https://arxiv.org/pdf/1907.07212). + S&P 2019. ### Algorithmic Fairness - Cynthia Dwork, Moritz Hardt, Toniann Pitassi, Omer Reingold, and Rich Zemel. diff --git a/website/docs/schedule/lectures.md b/website/docs/schedule/lectures.md index 54bd0bd..2cc4f55 100644 --- a/website/docs/schedule/lectures.md +++ b/website/docs/schedule/lectures.md @@ -7,34 +7,34 @@ 9/6 | Basic private mechanisms
**Reading:** AFDP 3.2-4 | 9/9 | Composition and closure properties
**Reading:** AFDP 3.5 | Signups 9/11 | What does differential privacy actually mean?
**Reading:** McSherry. [Lunchtime for Differential Privacy](https://github.com/frankmcsherry/blog/blob/master/posts/2016-08-16.md) | -9/13 | Paper presentations | HW1 Due +9/13 | Paper presentations: Differential privacy | HW1 Due |

**Adversarial Machine Learning**

| -9/16 | Overview and Basic attacks | HW2 Out -9/18 | More attacks | -9/20 | Paper presentations | -9/23 | Defense: Adversarial training | -9/25 | Defense: Certified defenses | -9/27 | Paper presentations | HW2 Due +9/16 | Overview and basic concepts | HW2 Out +9/18 | Paper presentations: Adversarial attacks | +9/20 | Paper presentations: ??? | +9/23 | Adversarial training | +9/25 | Paper presentations: Certified defenses | +9/27 | Paper presentations: ??? | HW2 Due |

**Applied Cryptography**

| 9/30 | Overview and basic constructions | HW3 Out -10/2 | Secure Multiparty Computation | -10/4 | Paper presentations | +10/2 | Paper presentations: Secure Multiparty Computation | +10/4 | Paper presentations: ??? | 10/7 | Homomorphic Encryption | -10/9 | Oblivious computing and side channels | -10/11 | Paper presentations | HW3 Due
MS1 Due +10/9 | Paper presentations: Oblivious computing and side channels | +10/11 | Paper presentations: ??? | HW3 Due
MS1 Due |

**Advanced Topic: Algorithmic Fairness**

| 10/14 | Overview and basic notions | -10/16 | Individual and group fairness | -10/18 | Paper presentations | -10/21 | Repairing fairness | -10/23 | Challenges in defining fairness | -10/25 | Paper presentations | +10/16 | Paper presentations: Individual and group fairness | +10/18 | Paper presentations: ??? | +10/21 | Challenges in defining fairness | +10/23 | Paper presentations: Repairing fairness | +10/25 | Paper presentations: ??? | |

**Advanced Topic: PL and Verification**

| 10/28 | Overview and basic notions | -10/30 | Programming languages for differential privacy | -11/1 | Paper presentations | -11/4 | Probabilistic programming languages | -11/6 | Verifying probabilistic programs | -11/8 | Paper presentations | MS2 Due +10/30 | Paper presentations: Probabilistic programming languages | +11/1 | Paper presentations: ??? | +11/4 | Programming languages for differential privacy | +11/6 | Paper presentations: Verifying probabilistic programs | +11/8 | Paper presentations: ??? | MS2 Due |

**No Lectures: Work on Projects**

| 12/11 (TBD) | Project Presentations |