Recent works have cast some light on the mystery of why deep nets fit any data and
generalize despite being very overparametrized. This paper analyzes training and …

One of the mysteries in the success of neural networks is randomly initialized first order
methods like gradient descent can achieve zero training loss even though the objective …

Infinite width limit has shed light on generalization and optimization aspects of deep learning
by establishing connections between neural networks and kernel methods. Despite their …

Understanding the uncertainty of a neural network's (NN) predictions is essential for many
applications. The Bayesian framework provides a principled approach to this, however …

Humans and animals routinely infer relations between different items or events and
generalize these relations to novel combinations of items. This allows them to respond …

A simple, flexible approach to creating expressive priors in Gaussian process (GP) models
makes new kernels from a combination of basic kernels, eg summing a periodic and linear …

Neural network models are known to reinforce hidden data biases, making them unreliable
and difficult to interpret. We seek to build models thatknow what they do not know'by …

I illustrate an approach that can be exploited for constructing neural networks that a priori
obey physical laws. We start with a simple single-layer neural network (NN) but refrain from …

A common problem in machine learning-based malware detection is that training data may
contain noisy labels and it is challenging to make the training data noise-free at a large …

Flexible models for probability distributions are an essential ingredient in many machine
learning tasks. We develop and investigate a new class of probability distributions, which we …