Scholarlink

First, we present the working (active) set method, which reduces the bound-constrained problem to a series of unconstrained problems solved by a direct solver. Then, we introduce the Polyak method, which implements a solution of auxiliary unconstrained problems by conjugate gradients. We show how to speed up the Polyak algorithm by “looking ahead” so it can use inexact solutions to auxiliary problems.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Active Sets with Finite Termination

  • 章节

  • 发表日期：

    2025年

查看全文
  • Zdeněk Dostál

摘要

First, we present the working (active) set method, which reduces the bound-constrained problem to a series of unconstrained problems solved by a direct solver. Then, we introduce the Polyak method, which implements a solution of auxiliary unconstrained problems by conjugate gradients. We show how to speed up the Polyak algorithm by “looking ahead” so it can use inexact solutions to auxiliary problems.