Syllabus and Schedule
April 26, 2022
14:00 - 14:15
Introduction
The need for privacy and the need for data
Private Computation Models
The Federated Model
Overview of privacy technologies
Cross-silo vs. cross-device setting
Analytics vs. Learning
Variants of federated computing
Limitations of Federated Computation
14:15 - 14:30
Technical Preliminaries
Definitions of privacy and security
Multi-party computation
Secure Aggregation
K-anonymity
Central and Local Differential Privacy
Event, Device and User privacy
14:30 - 15:00
Federated Algorithms
Secure Multi-Party Computation (MPC) for secure aggregation
DP frequency oracles and
histogramsAnalytics primitives: Sum, Mean,
Quantiles and Heavy HittersDistributed Noise Generation
Sample and Threshold privacy
15:00 - 15:15
​
Practical Considerations
Case studies of federation in practice
Current federated systems
Practical challenges
Data Characteristics
Building a federated stack
15:15 - 15:25
Open Problems
Cross-user analytics
Complex structured data
Longitudinal privacy and data streams
Defense against security threats
Real-time analytics
Proactive vs. reactive analytics
Federated learning and beyond
15:25 - 15:30
Q&A
The audience can feel free to ask any questions they have!