Adversarial attacks pose significant threats to deploying state-of-the-a...
In Semi-Supervised Semi-Private (SP) learning, the learner has access to...
Improving and guaranteeing the robustness of deep learning models has be...
Differential privacy is the de facto standard for protecting privacy in ...
Online learning, in the mistake bound model, is one of the most fundamen...
In supervised learning, it has been shown that label noise in the data c...
The vulnerability of machine learning models to spurious correlations ha...
Despite clear computational advantages in building robust neural network...
As machine learning algorithms are deployed on sensitive data in critica...
Recently, Wong et al. showed that adversarial training with single-step ...
Deep learning research has recently witnessed an impressively fast-paced...
We investigate two causes for adversarial vulnerability in deep neural
n...
Recent studies have shown that skeletonization (pruning parameters) of
n...
Miscalibration – a mismatch between a model's confidence and its correct...
Exciting new work on the generalization bounds for neural networks (NN) ...
Machine learning methods are widely used for a variety of prediction
pro...
A key feature of neural networks, particularly deep convolutional neural...
We present a class of algorithms capable of directly training deep neura...
Understanding the evolution of human society, as a complex adaptive syst...
We propose a generalization of neural network sequence models. Instead o...