Fair representation learning provides an effective way of enforcing fairness constraints without compromising utility for downstream users.

In our article, we introduce a Riemannian batch normalization (batch-norm) algorithm, which generalizes the one used in Euclidean nets.

Furthermore, with only 20% training data, we can match the top performance on CIFAR-10 and CIFAR-100. Finally, our method can generate high-fidelity images using only 100 images without pre-training, while being on par with existing transfer learning algorithms.

Compared with its rapid and widespread adoption, the theoretical understanding of why deep learning works so well has lagged significantly. This is particularly the case in the common setup of an overparameterized network, where the number of parameters in the network greatly exceeds the number of training examples and input dimension.

Motivated by this property, we propose a method to extend transformers to tree-structured data, enabling sequence-to-tree, tree-to-sequence, and tree-to-tree mappings.

Deep convolutional networks (DCNs) learn meaningful representations where data that share the same abstract characteristics are positioned closer and closer.

As shown in the latency results in experiments (Section 5.2), the extra variance-related computation only adds a

Bayesian neural networks and deep ensembles are principled approaches to estimate the predictive uncertainty of a deep learning model.

However, it suffers severe performance degradation.