Results 41 to 50 of about 12,020,063 (274)

The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods

open access: yesCoRR, 2021
The adaptive stochastic gradient descent (SGD) with momentum has been widely adopted in deep learning as well as convex optimization. In practice, the last iterate is commonly used as the final solution to make decisions. However, the available regret analysis and the setting of constant momentum parameters only guarantee the optimal convergence of the
Wei Tao 0002   +3 more
openaire   +4 more sources

Global convergence and optimality of the heavy ball method for non-convex optimization.

open access: yesCoRR, 2022
In this letter we revisit the famous heavy ball method and study its global convergence for a class of non-convex problems with sector-bounded gradient. We characterize the parameters that render the method globally convergent and yield the best $R$-convergence factor.
Valery A. Ugrinovskii   +2 more
openaire   +6 more sources

On the Global Convergence of Continuous–Time Stochastic Heavy–Ball Method for Nonconvex Optimization [PDF]

open access: yes2019 IEEE International Conference on Big Data (Big Data), 2019
We study the convergence behavior of the stochastic heavy-ball method with a small stepsize. Under a change of time scale, we approximate the discrete method by a stochastic differential equation that models small random perturbations of a coupled system of nonlinear oscillators. We rigorously show that the perturbed system converges to a local minimum
Wenqing Hu   +2 more
openaire   +2 more sources

Martian Golf Ball

open access: yes
This golf ball is designed to be used in Martian gravity. As Martian gravity is weaker than Earths, this golf ball feels heavy when held on Earth. It forms part of an ongoing project by Bethany Rigby on Golf and outer space, developing a speculative ...
Rigby, Bethany
core   +5 more sources

Synthesis and Characterization of Ground Biochar (GB) Reinforced Composites for Removal of Heavy Metal from Palm Oil Mill Effluent (POME)

open access: yesBioResources, 2023
Heavy metal contamination ruins the ecosystem and water quality. The adsorption method for heavy metal remediation is preferred because of its low cost and high efficiency.
Md Rezaur Rahman   +10 more
doaj   +2 more sources

Accelerated Convergence of Stochastic Heavy Ball Method under Anisotropic Gradient Noise

open access: yesCoRR, 2023
Published at ICLR ...
Rui Pan 0002   +3 more
openaire   +4 more sources

Facile solvent-free modified biochar for removal of mixed steroid hormones and heavy metals: isotherm and kinetic studies

open access: yesBMC Chemistry, 2023
Water contamination has become a global challenge to human survival. Non-biodegradable heavy metal cations and steroid hormones could accumulate in the human body and could result in serious health problems.
Sefiu Olaitan Amusat   +4 more
doaj   +1 more source

Applicability of carbothermic reduction for upgrading Sri Lankan ilmenite ores: towards converting ilmenite into synthetic rutile by mechanical activation

open access: yesBulletin of the National Research Centre, 2021
Background Ilmenite and rutile are naturally occurring titanium-bearing heavy minerals. Sri Lanka consists of ilmenite and rutile in placer deposits mainly along the northeast coast. The commercial value of rutile is higher than ilmenite.
T. Dilmi. U. Wijewardhana   +1 more
doaj   +1 more source

Analysis of a Two-Step Gradient Method with Two Momentum Parameters for Strongly Convex Unconstrained Optimization

open access: yesAlgorithms
The paper is devoted to the theoretical and numerical analysis of the two-step method, constructed as a modification of Polyak’s heavy ball method with the inclusion of an additional momentum parameter.
Gerasim V. Krivovichev   +1 more
doaj   +1 more source

Reliability-Based Robust Design Optimization with Fourth-Moment Method for Ball Bearing Wear

open access: yesLubricants
Ball bearings operating at low speeds and under heavy loads are susceptible to wear failure, leading to significant economic losses. The existing reliability-based robust design optimization method of the fourth-moment method has high accuracy and does ...
Yanzhong Wang   +4 more
doaj   +1 more source

Home - About - Disclaimer - Privacy