Overview
Explore the critical concepts of privacy and fairness in computing through this 45-minute ACM conference talk. Delve into local and central differential privacy, stochastic convex optimization, and new algorithms for addressing privacy concerns. Examine the nonsmooth case, identify gaps in current approaches, and understand why interaction is necessary in privacy-preserving systems. Investigate various learning settings and their implications for privacy. Discover practical applications, including committee selection and stable committees, and engage in a thought-provoking question-and-answer session to deepen your understanding of these crucial topics in modern computing.
Syllabus
Intro
Local and Central Differential Privacy
QA
Stochastic convex optimization
New algorithms
Nonsmooth case
Whats missing
Interaction is necessary
Learning settings
Questions
Committee Selection
Applications
Stable Committees
Question Time
Taught by
Association for Computing Machinery (ACM)