Results 61 to 70 of about 1,415,554 (169)

Gradient Coding

CoRR, 2016
We propose a novel coding theoretic framework for mitigating stragglers in distributed learning. We show how carefully replicating data blocks and coding across gradients can provide tolerance to failures and stragglers for Synchronous Gradient Descent. We implement our schemes in python (using MPI) to run on Amazon EC2, and show how we compare against
Rashish Tandon, Qi Lei, Alexandros G. Dimakis, Nikos Karampatziakis   +3 more
openaire   +2 more sources

Gradient networks

Journal of Physics A: Mathematical and Theoretical, 2008
We define gradient networks as directed graphs formed by local gradients of a scalar field distributed on the nodes of a substrate network G. We derive an exact expression for the in-degree distribution of the gradient network when the substrate is a binomial (Erdos-Renyi) random graph, G(N,p).
Toroczkai, Zoltan, Kozma, Balazs, Bassler, Kevin E., Hengartner, N. W., Korniss, G.   +4 more
openaire   +2 more sources

On Gradient Flows

Journal of Differential Equations, 1998
The purpose of this paper is to contribute to the theory of existence of solutions to ordinary differential equations in \(\mathbb{R}^n\), when the right hand side of the equation is not necessarily continuous. More precisely the authors study existence and uniqueness of solutions for the equation \(x'= \nabla u(x)\) when \(u\) is not necessarily ...
CELLINA, ARRIGO, Vornicescu, M.
openaire   +1 more source

Effet de l'activité des insectes pollinisateurs sur la pollinisation et le rendement du tournesol de consommation

Oilseeds and fats, crops and lipids, 2017
La pollinisation entomophile du tournesol est souvent mentionnée comme un facteur contribuant au rendement et à la qualité de cette culture. Alors que les rendements stagnent depuis une trentaine d'années en France, l'amélioration de la pollinisation des
Fougeroux André, Leylavergne Solenne, Guillemard Vincent, Geist Olivier, Gary Pauline, Cenier Charlotte, Caumes-Sudre Edith, Senechal Christopher, Vaissière Bernard   +8 more
doaj   +1 more source

Gradient Networks

IEEE Transactions on Signal Processing
Directly parameterizing and learning gradients of functions has widespread significance, with specific applications in inverse problems, generative modeling, and optimal transport. This paper introduces gradient networks (GradNets): novel neural network architectures that parameterize gradients of various function classes.
Shreyas Chaudhari, Srinivasa Pranav, José M. F. Moura   +2 more
openaire   +2 more sources

Comparative analysis of asbestos body and fiber content in formalin-fixed vs. paraffin-embedded lung tissue

Frontiers in Public Health
IntroductionAsbestos body and fiber burdens may be determined using different preparations of lung tissue. Paraffin-embedded tissue requires more complex steps than formalin-fixed tissue.
Barbara K. Kuhn, Bart Vrugt, Bart Vrugt, Alex Low, Julie E. Goodman, Richard Attanoos   +5 more
doaj   +1 more source

Passive seismic data processing methods to identify contrast intrasalt interlayers in the geological section of the Astrakhan Arch

Georesursy
The article examines the problem of processing microseismic noise (MN) to identify and evaluate occurrence depth of contrasting geological objects – intersalt interlayers with a potentially high formation pressure.
E. V. Biryaltsev, A. A. Vikhoreva, V. A. Zakharchuk, A. Yu. Komarov, V. V. Pykhalov, O. V. Tinakin   +5 more
doaj   +1 more source

Nonlinear Analysis of the Effects of Socioeconomic, Demographic, and Technological Factors on the Number of Fatal Traffic Accidents

Safety
This study explores the complex connections among various socioeconomic, demographic, and technological aspects and their impact on fatal traffic accidents.
Nassim Sohaee, Shahram Bohluli
doaj   +1 more source

Source Tracing of PM2.5 in a Metropolitan Area Using a Low-Cost Air Quality Monitoring Network: Case Study of Denver, Colorado, USA

Atmosphere
Air quality assessments often require source apportioning of the air pollutants observed at the receptor site. Conventional source apportionment models are subject to high uncertainties due to the lack of accurate emission profiles of all the ...
Nima Afshar-Mohajer, Mirella Shaban
doaj   +1 more source

Function Analysis of the Euclidean Distance between Probability Distributions

Entropy, 2018
Minimization of the Euclidean distance between output distribution and Dirac delta functions as a performance criterion is known to match the distribution of system output with delta functions.
Namyong Kim
doaj   +1 more source