Gradients Look Alike: Sensitivity is Often Overestimated in DP-SGD.

Anvith Thudi Hengrui Jia Casey Meehan Ilia Shumailov Nicolas Papernot

Published in: CoRR (2023)

Keyphrases

dynamic programming
sensitivity analysis
neural network
high sensitivity
database
machine learning
genetic algorithm
stochastic gradient descent
reinforcement learning
support vector
recommender systems
cost function
gradient orientation