Suchergebnisse - "First order algorithms"

Randomized first order algorithms with applications to ℓ 1-minimization von Juditsky, Anatoli, Kılınç Karzan, Fatma, Nemirovski, Arkadi

Volltext

Tradeoffs Between Convergence Rate and Noise Amplification for Momentum-Based Accelerated Optimization Algorithms von Mohammadi, Hesameddin, Razaviyayn, Meisam, Jovanovic, Mihailo R.

“… In this article, we study momentum-based first-order optimization algorithms in which the iterations utilize information from the two previous steps and are subject to an additive white noise …”
Volltext

S-NEAR-DGD: A Flexible Distributed Stochastic Gradient Method for Inexact Communication von Iakovidou, Charikleia, Wei, Ermin

“… Our method is based on a class of flexible, distributed first-order algorithms that allow for the tradeoff of computation and communication to best accommodate the application setting …”
Volltext

Robustness of First- and Second-Order Consensus Algorithms for a Noisy Scale-Free Small-World Koch Network von Yi, Yuhao, Zhang, Zhongzhi, Shan, Liren, Chen, Guanrong

“… We focus on three cases of consensus schemes: (1) first-order leaderless algorithm; (2) first-order algorithm with a single leader; and (3 …”
Volltext

Finding Second-Order Stationary Points in Constrained Minimization: A Feasible Direction Approach von Hallak, Nadav, Teboulle, Marc

“… The first-order step is a generic closed map algorithm, which can be chosen from a variety of first-order algorithms, making it adjustable to the given problem …”
Volltext

Trilevel and multilevel optimization using monotone operator theory von Shafiei, Allahkaram, Kungurtsev, Vyacheslav, Marecek, Jakub

“…  Based on fixed-point theory and related arguments, we present a natural first-order algorithm and analyze its convergence and rates of convergence in several regimes …”
Volltext

Optimal First-Order Algorithms as a Function of Inequalities von Park, Chanwoo, Ryu, Ernest K

“… Specifically, we restrict convergence analyses of algorithms to use a prespecified subset of inequalities, rather than utilizing all true inequalities, and find the optimal algorithm subject to this restriction …”
Volltext

First-Order Fast Algorithm for Structurally Optimal Multi-Group Multicast Beamforming in Large-Scale Systems von Zhang, Chong, Dong, Min, Liang, Ben

“… Based on the optimal multicast beamforming structure, we propose a fast first-order algorithm to obtain the beamforming solution …”
Volltext

Trilevel and Multilevel Optimization using Monotone Operator Theory von Shafiei, Allahkaram, Kungurtsev, Vyacheslav, Marecek, Jakub

“… ~Based on fixed-point theory and related arguments, we present a natural first-order algorithm and analyze its convergence and rates of convergence in several regimes …”
Volltext

A likelihood-based approach for multivariate categorical response regression in high dimensions von Molstad, Aaron J, Rothman, Adam J

“… both the marginal distributions and log odds ratios. To compute our estimator, we propose an efficient first order algorithm which we extend to settings where some subjects have only one response variable measured, i.e …”
Volltext

Aggregating regular norms von Juditsky, Anatoli, Nemirovski, Arkadi

“… ) high-dimensional convex geometry and probability in Banach spaces [0.9.12.13.15], and in 2) design of proximal first-order algorithms for large-scale convex optimization with dimension-independent, or nearly so, complexity …”
Volltext

Robustly Stable Accelerated Momentum Methods With A Near-Optimal L2 Gain and \(H_\infty\) Performance von Gurbuzbalaban, Mert

“… We study the trade-offs between the convergence rate and robustness to gradient errors when designing the parameters of a first-order algorithm …”
Volltext

Optimal Sparse \(H_\infty\) Controller Design for Networked Control Systems von Yang, Zhaohua, Wang, Pengyu, Zhang, Haishan, Jia, Shiyue, Yang, Nachuan, Zhong, Yuxing, Shi, Ling

“… However, the design of optimal sparse \(H_\infty\) controllers remains an open and challenging problem due to its non-convexity, and we cannot design a first-order algorithm to analyze since we lack an analytical expression for a given controller …”
Volltext

S-NEAR-DGD: A Flexible Distributed Stochastic Gradient Method for Inexact Communication von Iakovidou, Charikleia, Wei, Ermin

“… Our method is based on a class of flexible, distributed first order algorithms that allow for the trade-off of computation and communication to best accommodate the application setting …”
Volltext

Entropic Risk-Averse Generalized Momentum Methods von Can, Bugra, Gürbüzbalaban, Mert

“… In the context of first-order algorithms subject to random gradient noise, we study the trade-offs between the convergence rate …”
Volltext

Fast First-Order Methods for Monotone Strongly DR-Submodular Maximization von Sadeghi, Omid, Fazel, Maryam

“… ) property, which implies that they are concave along non-negative directions. Existing works have studied monotone continuous DR-submodular maximization subject …”
Volltext

COCO Denoiser: Using Co-Coercivity for Variance Reduction in Stochastic Convex Optimization von Madeira, Manuel, Negrinho, Renato, Xavier, João, Aguiar, Pedro M Q

“… Our method, named COCO denoiser, is the joint maximum likelihood estimator of multiple function gradients from their noisy observations, subject to co-coercivity constraints …”
Volltext

On the effect of perturbations in first-order optimization methods with inertia and Hessian driven damping von Attouch, Hedy, Fadili, Jalal, Kungurtsev, Vyacheslav

“… Second-order continuous-time dissipative dynamical systems with viscous and Hessian driven damping have inspired effective first-order algorithms for solving convex optimization problems …”
Volltext

Joint Fronthaul and Multicast Beamforming for Ultra-Dense C-RANs von Tan, Fangqing, Wu, Peiran, Jiang, Ming, Xia, Minghua

“… algorithm with low computational complexity is developed to find the optimal solution. Numerical results demonstrate that the proposed first-order algorithm can achieve …”
Volltext

Primal-Dual Methods for Saddle-Point Problems with Applications to Decentralized Constrained Convex Optimization von Yazdandoost Hamedani, Erfan

“… Saddle-point (SP) problems form an important class of computational problems with the aim of minimizing a function over one variable while maximizing over the …”
Volltext