
Ravi Kumar
Google
[intermediate/advanced] Differential Privacy
Summary
This course will introduce differential privacy (DP) as an important tool for private analytics and machine learning.
Syllabus
- DP: Motivation; Definition; Properties
- Basic Mechanisms: Laplace; Gaussian; Exponential
- Private Analytics: Histograms, Quantiles
- Private ML: SGD; Hyperparameters
- Local/Shuffle models: Properties; Algorithms
- Advanced Methods: Synthetic data; Matrix mechanism
References
Pre-requisites
Basic probability. Basic machine learning.
Short bio
Ravi Kumar has been a research scientist at Google since 2012. Prior to this, he was at the IBM Almaden Research Center and at Yahoo! Research. His interests include algorithms for massive data, ML/privacy, and the theory of computation.