首页 | 本学科首页   官方微博 | 高级检索  
     


The radar method: an effective line search for piecewise linear concave functions
Authors:C. Beltran-Royo
Affiliation:(1) DEIO, Universidad Rey Juan Carlos, Mostoles, Madrid, Spain
Abstract:The maximization of one-dimensional piecewise linear concave (OPLC) functions arises in the line search associated with the maximization of piecewise linear concave functions (e.g. Kelley cutting plane method). The OPLC line search is usually done by the next-break-point method, where one goes from break point to break point up to the optimum. If the number of break points is large this method will be computationally expensive. One can also use some classical derivative-free line search method as for example the golden section method. Such methods do not take advantage of the OPLC geometry. As an alternative, we propose an improved version of the so-called radar method, which maximizes an OPLC function by maximizing successive outer approximations. We prove superlinear and finite convergence of the radar method. Furthermore, our computational test shows that the radar method is highly effective independently from the number of break points.
Keywords:Piecewise linear concave function  Line search  Radar method  Next-break-point method  Golden section method  Kelley cutting plane method
本文献已被 SpringerLink 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号