Privacy Amplification by Iteration for ADMM with (Strongly) Convex Objective Functions.
T.-H. Hubert ChanHao XieMengshi ZhaoPublished in: AAAI (2024)
Keyphrases
- objective function
- convex optimization
- alternating direction method of multipliers
- convex functions
- quasiconvex
- total variation
- privacy preserving
- piecewise linear
- personal information
- convex hull
- optimization problems
- privacy concerns
- multi objective
- global optimality
- line search
- newton raphson
- security issues
- private information
- cost function
- privacy protection
- linear programming
- global optimum
- quadratic program
- optimal solution
- convex sets
- interior point methods
- feasible solution
- personal data
- privacy preservation
- image denoising
- linear program
- private data
- statistical databases
- privacy preserving data mining
- denoising
- low rank
- augmented lagrangian method
- convex relaxation
- norm minimization
- data mining
- augmented lagrangian
- risk minimization
- sensitive data
- globally optimal
- sensitive information