Timezone: »

Differentially private training of neural networks with Langevin dynamics for calibrated predictive uncertainty
Moritz Knolle · Alexander Ziller · Dmitrii Usynin · Rickmer Braren · Marcus Makowski · Daniel Rueckert · Georgios Kaissis

We show that differentially private stochastic gradient descent (DP-SGD) can yield poorly calibrated, overconfident deep learning models. This represents a serious issue for safety-critical applications, e.g. in medical diagnosis. We highlight and exploit parallels between stochastic gradient Langevin dynamics, a scalable Bayesian inference technique for training deep neural networks,and DP-SGD, in order to train differentially private, Bayesian neural networks with minor adjustments to the original (DP-SGD) algorithm.Our approach provides considerably more reliable uncertainty estimates than DP-SGD, as demonstrated empirically by a reduction in expected calibration error (MNIST∼5-fold, Pediatric Pneumonia Dataset∼2-fold).

Author Information

Moritz Knolle (Technical University Munich)
Alexander Ziller (Technische Universität München)
Dmitrii Usynin (Imperial College London / TU Munich)
Rickmer Braren
Marcus Makowski
Daniel Rueckert (Imperial College London)
Georgios Kaissis (Technical University Munich)

More from the Same Authors