A non-monotone trust region method
Journal Title: Scholars Bulletin - Year 2016, Vol 2, Issue 1
Abstract
Abstract: In this paper, we propose a non-monotone trust region algorithm based on conic model for unconstrained optimization problems. The method combines a non-monotone technique and a trust region method. Different from the usual trust region methods, the sub-problem of our new algorithm is the conic minimization sub-problem. The theoretical analysis indicates that the new method has the global convergence under some reasonable conditions. Keywords: non-monotone technique; trust region method; line search; unconstrained optimization; conic model; global convergence.
Authors and Affiliations
Liran Yang, Qinghua Zhou
Keywords
Related Articles
- EP ID EP379069
- DOI -
- Views 120
- Downloads 0
An Analysis of Discourse Markers in Football commentaries
Abstract:The importance of discourse markers lies in the theoretical questions they raise about the nature of discourse and the relationship between linguistic meaning and context. Schiffrin described discourse markers a...
The current situation and prospects of remaining oil distribution
Abstract:The study of remaining oil distribution is always an important and challenging work, accurate evaluation of remaining oil. And the degree of the middle and later periods of the oilfield development adjustment ta...
Non-Monotone Wedge Trust-Region Method for Derivative-Free Unconstrained Optimization
Abstract:We propose a non-monotone wedge trust region method for derivative-free optimization. Wedge trust region method based on traditional trust region is designed for derivative-free problems, and the non-monotone st...
Influence of Graphite Waste Incorporation on the Properties of Epoxy Matrix Composites
Abstract:The addition of filler materials to a polymer is a common practice to improve strength, stiffness, toughness, hardness, conductivity, wear resistance, as well as reducing the processing cost significantly. In th...
A non-monotone trust region method
Abstract: In this paper, we propose a non-monotone trust region algorithm based on conic model for unconstrained optimization problems. The method combines a non-monotone technique and a trust region method. Different fr...