A Stochastic Trust Region Method for Unconstrained Optimization Problems. (3rd February 2019)

• Record Type: Journal Article
• Authors: Li, Ningning; Xue, Dan; Sun, Wenyu; Wang, Jing
• Other Names: Manfredi Paolo Academic Editor.
• Is Part Of: Mathematical problems in engineering. Volume 2019(2019)
• Publication Date: 2019-02-03
• Subjects: Engineering mathematics -- Periodicals; 510.2462
• DOI: 10.1155/2019/8095054
• Languages: English
• ISSNs: 1024-123X

Abstract

Abstract : In this paper, a stochastic trust region method is proposed to solve unconstrained minimization problems with stochastic objectives. Particularly, this method can be used to deal with nonconvex problems. At each iteration, we construct a quadratic model of the objective function. In the model, stochastic gradient is used to take the place of deterministic gradient for both the determination of descent directions and the approximation of the Hessians of the objective function. The behavior and the convergence properties of the proposed method are discussed under some reasonable conditions. Some preliminary numerical results show that our method is potentially efficient.