A Proof of Exact Convergence Rate of Gradient Descent. Part II. Performance Criterion $(f(x_N)-f_*)/\|x_0-x_*\|^2$

Autor:	Kim, Jungbin
Rok vydání:	2024
Předmět:	Mathematics - Optimization and Control
Druh dokumentu:	Working Paper
Popis:	We prove the exact worst-case convergence rate of gradient descent for smooth strongly convex optimization, with respect to the performance criterion $(f(x_N)-f_*)/\Vert x_0-x_*\Vert^2$. This rate was previously conjectured in [DT14; THG17].
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2412.04427 Zobrazit plný text záznamu View this record from Arxiv