New Primitives for Actively-Secure MPC over Rings with Applications to Private Machine Learning
Offered By: IEEE via YouTube
Course Description
Overview
Explore a conference talk on new primitives for actively-secure Multi-Party Computation (MPC) over rings, with applications to private machine learning. Delve into the SPDZ2k protocol presented at CRYPTO 2018, which operates over rings instead of fields, potentially allowing for more efficient implementations using native CPU operations. Examine the implementation of this scheme and the development of new efficient protocols for equality testing, comparison, and truncation over rings. Discover how these operations apply to machine learning, particularly in oblivious algorithms for decision tree and support vector machine (SVM) evaluation. Learn about the benchmarks that demonstrate the superior performance of these ring-based approaches compared to their field-based counterparts.
Syllabus
Intro
What is MPC
Protocol
Primitives
Generation
Benchmarks
Questions
Taught by
IEEE Symposium on Security and Privacy
Tags
Related Courses
Applied CryptographyUniversity of Virginia via Udacity Cryptography II
Stanford University via Coursera Coding the Matrix: Linear Algebra through Computer Science Applications
Brown University via Coursera Cryptography I
Stanford University via Coursera Unpredictable? Randomness, Chance and Free Will
National University of Singapore via Coursera