Faster Rates of Convergence to Stationary Points in Differentially Private Optimization.
Raman AroraRaef BassilyTomás GonzálezCristóbal GuzmánMichael MenartEnayat UllahPublished in: CoRR (2022)
Keyphrases
- differentially private
- rates of convergence
- stationary points
- constrained optimization
- nonlinear programming
- differential privacy
- optimization problems
- regression function
- objective function
- mathematical programming
- fixed point
- machine learning
- quadratic programming
- optimization method
- expectation maximization
- convergence analysis
- semi supervised
- active learning
- lower bound
- training data