Prioritized training on points that are learnable, worth learning, and not yet learnt

S Mindermann, JM Brauner… - International …, 2022‏ - proceedings.mlr.press
Training on web-scale data can take months. But much computation and time is wasted on
redundant and noisy points that are already learnt or not learnable. To accelerate training …

Deep sets

M Zaheer, S Kottur, S Ravanbakhsh… - Advances in neural …, 2017‏ - proceedings.neurips.cc
We study the problem of designing models for machine learning tasks defined on sets. In
contrast to the traditional approach of operating on fixed dimensional vectors, we consider …

Training deep spiking neural networks using backpropagation

JH Lee, T Delbruck, M Pfeiffer - Frontiers in neuroscience, 2016‏ - frontiersin.org
Deep spiking neural networks (SNNs) hold the potential for improving the latency and
energy efficiency of deep neural networks through data-driven event-based computation …

Generalized byzantine-tolerant sgd

C **e, O Koyejo, I Gupta - arxiv preprint arxiv:1802.10116, 2018‏ - arxiv.org
We propose three new robust aggregation rules for distributed synchronous Stochastic
Gradient Descent~(SGD) under a general Byzantine failure model. The attackers can …

One network to solve them all--solving linear inverse problems using deep projection models

JH Rick Chang, CL Li, B Poczos… - Proceedings of the …, 2017‏ - openaccess.thecvf.com
While deep learning methods have achieved state-of-the-art performance in many
challenging inverse problems like image inpainting and super-resolution, they invariably …

Adaptive data augmentation for image classification

A Fawzi, H Samulowitz, D Turaga… - 2016 IEEE international …, 2016‏ - ieeexplore.ieee.org
Data augmentation is the process of generating samples by transforming training data, with
the target of improving the accuracy and robustness of classifiers. In this paper, we propose …