Security and Privacy for Distributed Optimization & Distributed Machine Learning
Nitin H. Vaidya · 2021
The tutorial will include an introduction to (i) distributed optimization and distributed machine learning, (ii) security or fault-tolerance for distributed optimization and learning, and (iii) privacy in distributed optimization and learning. The presentation will cover the basic principles, and some representative solutions. Server-based and peer-to-peer solutions will be discussed. In particular, Byzantine fault-tolerant algorithms for distributed optimization and learning will be discussed. Privacy mechanisms to be discussed include differential privacy and its variations for systems based on multiple servers.