LATENT reports 91 to 96 percent accuracy at epsilon=0.5 for split-CNN local differential privacy, but its privacy proof is incomplete and the mechanism as specified is not epsilon-LDP.
Machine learning models that remember too much,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Local Differential Privacy for Deep Learning
LATENT reports 91 to 96 percent accuracy at epsilon=0.5 for split-CNN local differential privacy, but its privacy proof is incomplete and the mechanism as specified is not epsilon-LDP.