Final Project

  • The guidelines and some recommended topics can be found here.

  • You have to choose your project by November 20.

  • Each student has to make a presentation, scheduled for the last lecture, and submit a report up to 5 pages by December 15. The LaTex format for the report can be found here.

The schedule of the final presentations:

Presenter time (Dec 6) Title
Sarthak 10.30 - 11.00 Differentially Private Empirical Risk Minimization
Mohammad 11.00 - 11.30 Semi-Supervised Knowledge Transfer for Deep Learning from Private Training Data
Jason 11.30 - 12.00 Node and Edge Differential Privacy for Graph Laplacian Spectra: Mechanisms and Scaling Laws
Yinying 12.00 - 12.30 Collecting Telemetry Data Privately
Alireza 12.30 - 1.00 Black-Box Differential Privacy for Interactive ML
Yuanqi 1.00 - 1.30 On the Privacy Risks of Algorithmic Recourse
Narges 1.30 - 2.00 Concentrated Differentially Private Gradient Descent with Adaptive per-Iteration Privacy Budget
Catherine 2.00 - 2.30 Attacks on Deidentification’s Defenses