arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.12464math.OC
收藏

黎曼流形上的一种自适应三次正则化拟牛顿方法

An Adaptive Cubic Regularization quasi-Newton Method on Riemannian Manifolds

Mauricio S. Louzeiro, Gilson N. Silva, Jinyun Yuan, Daoping Zhang

AI总结:

本文提出一种自适应三次正则化拟牛顿方法,用于求解黎曼无约束非凸优化问题,在目标函数梯度和 Hessian 矩阵未知时依然适用,并具备良好的迭代复杂度理论保证与数值表现。

摘要:

设计了一种带三次正则化的拟牛顿方法,用于求解黎曼无约束非凸优化问题。该算法是完全自适应的,对于给定的 $ε_g$,至多需要 ${\cal O} (ε_g^{-3/2})$ 次迭代即可使梯度小于 $ε_g$,且至多需要 $\mathcal O(\max\{ ε_g^{-\frac{3}{2}}, ε_H^{-3} \})$ 次迭代以达到二阶驻点。值得注意的是,即使目标函数的梯度和 Hessian 矩阵未知,该算法依然适用。通过使用前向有限差分近似梯度和 Hessian 进行了数值实验,以验证理论结果并进行数值比较。

英文摘要:

A quasi-Newton method with cubic regularization is designed for solving Riemannian unconstrained nonconvex optimization problems. The proposed algorithm is fully adaptive with at most ${\cal O} (ε_g^{-3/2})$ iterations to achieve a gradient smaller than $ε_g$ for given $ε_g$, and at most $\mathcal O(\max\{ ε_g^{-\frac{3}{2}}, ε_H^{-3} \})$ iterations to reach a second-order stationary point respectively. Notably, the proposed algorithm remains applicable even in cases of the gradient and Hessian of the objective function unknown. Numerical experiments are performed with gradient and Hessian being approximated by forward finite-differences to illustrate the theoretical results and numerical comparison.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑