Globally convergent limited memory bundle method for large-scale nonsmooth optimization

• 成果类型: Article
• 署名作者: Haarala, Napsu; Miettinen, Kaisa; Makela, Marko M.
• 署名单位: University of Witwatersrand; Aalto University; University of Jyvaskyla
• 刊物名称: MATHEMATICAL PROGRAMMING
• ISSN/ISSBN: 0025-5610
• DOI: 10.1007/s10107-006-0728-2
• 发表日期: 2007
• 页码: 181-205
• 关键词: variable-metric method; quasi-newton matrices; power management

摘要

Many practical optimization problems involve nonsmooth (that is, not necessarily differentiable) functions of thousands of variables. In the paper [Haarala, Miettinen, Makela, Optimization Methods and Software, 19, (2004), pp. 673-692] we have described an efficient method for large-scale nonsmooth optimization. In this paper, we introduce a new variant of this method and prove its global convergence for locally Lipschitz continuous objective functions, which are not necessarily differentiable or convex. In addition, we give some encouraging results from numerical experiments.