Randomized subspace regularized newton method for unconstrained non-convex optimization

T Fuji, PL Poirion, A Takeda - arXiv preprint arXiv:2209.04170, 2022 - arxiv.org
While there already exist randomized subspace Newton methods that restrict the search
direction to a random subspace for a convex function, we propose a randomized subspace …

A globally convergent gradient method with momentum

M Lapucci, G Liuzzi, S Lucidi, M Sciandrone - arXiv preprint arXiv …, 2024 - arxiv.org
In this work, we consider smooth unconstrained optimization problems and we deal with the
class of gradient methods with momentum, ie, descent algorithms where the search direction …

Subspace Quasi-Newton Method with Gradient Approximation

T Miyaishi, R Nozawa, PL Poirion, A Takeda - arXiv preprint arXiv …, 2024 - arxiv.org
In recent years, various subspace algorithms have been developed to handle large-scale
optimization problems. Although existing subspace Newton methods require fewer iterations …

Limited-memory Common-directions Method With Subsampled Newton Directions for Large-scale Linear Classification

JN Yen, CJ Lin - 2021 IEEE International Conference on Data …, 2021 - ieeexplore.ieee.org
The common-directions method is an optimization method recently proposed to utilize
second-order information. It is especially efficient on large-scale linear classification …

A Novel Fast Exact Subproblem Solver for Stochastic Quasi-Newton Cubic Regularized Optimization

J Forristal, J Griffin, W Zhou, S Yektamaram - arXiv preprint arXiv …, 2022 - arxiv.org
In this work we describe an Adaptive Regularization using Cubics (ARC) method for large-
scale nonconvex unconstrained optimization using Limited-memory Quasi-Newton (LQN) …