Publication

Stochastic distributed learning with gradient quantization and double-variance reduction

Publications associées (62)

Graph Chatbot

Chattez avec Graph Search

Posez n’importe quelle question sur les cours, conférences, exercices, recherches, actualités, etc. de l’EPFL ou essayez les exemples de questions ci-dessous.

AVERTISSEMENT : Le chatbot Graph n'est pas programmé pour fournir des réponses explicites ou catégoriques à vos questions. Il transforme plutôt vos questions en demandes API qui sont distribuées aux différents services informatiques officiellement administrés par l'EPFL. Son but est uniquement de collecter et de recommander des références pertinentes à des contenus que vous pouvez explorer pour vous aider à répondre à vos questions.

Connectez-vous pour utiliser Chat avec Graph Search

Stochastic Zeroth-Order Optimisation Algorithms with Variance Reduction

Ahmad Ajalloeian

Introduction of optimisation problems in which the objective function is black box or obtaining the gradient is infeasible, has recently raised interest in zeroth-order optimisation methods. As an example finding adversarial examples for Deep Learning mode ...

2019

We propose a class of novel variance-reduced stochastic conditional gradient methods. By adopting the recent stochastic path-integrated differential estimator technique (SPIDER) of Fang et al. (2018) for the classical Frank-Wolfe (FW) method, we introduce ...

2019

Lossy compression of quality scores in differential gene expression: A first assessment and impact analysis

Ana Angelica Hernandez Lopez

High-throughput sequencing of RNA molecules has enabled the quantitative analysis of gene expression at the expense of storage space and processing power. To alleviate these prob- lems, lossy compression methods of the quality scores associated to RNA sequ ...

2018

We develop a primal-dual convex minimization framework to solve a class of stochastic convex three-composite problem with a linear operator. We consider the cases where the problem is both convex and strongly convex and analyze the convergence of the propo ...

JMLR: W&CP2018

, , , , , ,

We experimentally demonstrate multiple Kerr-frequency-comb generation using different lines from another Kerr comb located up to a 50 km distance. The master and generated slave combs are mutually coherent and have a small variance of frequency error. ...

IEEE2018

Differential Gene Expression with Lossy Compression of Quality Scores in RNA-Seq Data

Marco Mattavelli, Claudio Paolo Alberti, Ana Angelica Hernandez Lopez

High-Throughput sequencing of RNA molecules has enabled the quantitative analysis of the expression of genes at the expense of storage space and processing power. To help alleviate these problems, lossy compression methods of the quality scores associated ...

2017

Fully Quantized Distributed Gradient Descent

n major distributed optimization system, the main bottleneck is often the communication between the different machines. To reduce the time dedicated to communications, some heuristics have been developed to reduce the precision of the messages sent and hav ...

2017

Non-convex, non-local functionals converging to the total variation

Hoài-Minh Nguyên

We present new results concerning the approximation of the total variation, integral(Omega)vertical bar del u vertical bar, of a function u by non-local, non-convex functionals of the form Lambda delta(u) = integral(Omega)integral(Omega)delta phi(vertical ...

Elsevier2017

Separation with restricted families of sets

János Pach, Gábor Tardos, Márton Naszódi, Zsolt Langi

Given a finite n-element set X, a family of subsets F subset of 2(X) is said to separate X if any two elements of X are separated by at least one member of F. It is shown that if vertical bar F vertical bar > 2(n-1), then one can select vertical bar log n ...

Academic Press Inc Elsevier Science2016

Broadband Uniform Wavelength Conversion and Time Compression of WDM Channels

Camille Sophie Brès, Mohammad Amin Shoaie

A scheme based on 2-pump OPA is proposed for uniform wavelength conversion and optimized compression. We show 4.7-fold compression over 32 nm range resulting in Gaussian pulses from sinusoidal modulation and enabling simultaneous compression of WDM channel ...

2015