Differentially Private Learning Needs Hidden State (Or Much Faster Convergence).
Jiayuan YeReza ShokriPublished in: NeurIPS (2022)
Keyphrases
- hidden state
- faster convergence
- differentially private
- learning algorithm
- reinforcement learning
- social networks
- hidden markov models
- convergence speed
- image segmentation
- computational complexity
- evolutionary algorithm
- wavelet transform
- markov random field
- simulated annealing
- particle swarm optimization
- domain independent
- markov models
- differential privacy
- neural network