Results 31 to 40 of about 2,575,606 (330)
Energies, 2019 In annealing, steel coils should be heated and consequently cooled according to the technological prescription defined for the annealed type of steel. It is appropriate to develop the systems and methods for estimation of the steel coil inner temperature
Milan Durdán +3 more
doaj +1 more source
Gradient Method with Step Adaptation
MathematicsThe paper solves the problem of constructing step adjustment algorithms for a gradient method based on the principle of the steepest descent. The expansion of the step adjustment principle, its formalization and parameterization led the researchers to ...
Vladimir Krutikov +4 more
doaj +1 more source
On the linear convergence of the stochastic gradient method with constant step-size [PDF]
, 2018 The strong growth condition (SGC) is known to be a sufficient condition for linear convergence of the stochastic gradient method using a constant step-size $\gamma$ (SGM-CS).
Cevher, Volkan, Vu, Bang Cong
core +2 more sources
Curvature-aided Incremental Aggregated Gradient Method
, 2017 We propose a new algorithm for finite sum optimization which we call the curvature-aided incremental aggregated gradient (CIAG) method. Motivated by the problem of training a classifier for a d-dimensional problem, where the number of training data is $m$
Nedic, Angelia +3 more
core +1 more source
Mathematics, 2023 This paper considers a model for the accumulation of mutations in a population of mice with a weakened function of polymerases responsible for correcting DNA copying errors during cell division.
Raul Argun +4 more
doaj +1 more source
The q-gradient method for global optimization
, 2013 The q-gradient is an extension of the classical gradient vector based on the concept of Jackson's derivative. Here we introduce a preliminary version of the q-gradient method for unconstrained global optimization. The main idea behind our approach is the
Galski, Roberto L. +2 more
core +1 more source
Компьютерная оптика, 2023 We consider a gradient method for calculating cascaded diffractive optical elements (DOEs) consisting of several sequentially placed phase DOEs. Using the unitarity property of the operator describing the light propagation through the cascaded DOE, we ...
D.V. Soshnikov +2 more
doaj +1 more source
Mathematical Model for Establishing the Time-Dependent Behavior of Rocks by the Gradient Method
Applied Sciences, 2022 In the underground activity domain, most problems related to mining pressure and mining stability need to be solved by taking into account the time behavior of rocks through an approach of the interaction amidst the rock massif, support system, time ...
Mihaela Toderas
doaj +1 more source
Reducing computational costs in deep learning on almost linearly separable training data [PDF]
Компьютерная оптика, 2020 Previous research in deep learning indicates that iterations of the gradient descent, over separable data converge toward the L2 maximum margin solution.
Ilona Kulikovskikh
doaj +1 more source
Background field method in the gradient flow
, 2015 In perturbative consideration of the Yang--Mills gradient flow, it is useful to introduce a gauge non-covariant term ("gauge-fixing term") to the flow equation that gives rise to a Gaussian damping factor also for gauge degrees of freedom. In the present
Suzuki, Hiroshi
core +2 more sources