The Bulletin of Irkutsk State University. Series Mathematics, 2015, Volume 11, Pages 54–68
Using of modified simplex imbeddings method for solving special class of convex non-differentiable optimization problems
A. V. Kolosnitcyn
Melentiev Energy Systems Institute SB RAS, 130, Lermontov st., Irkutsk, 664033
In this paper it is considered the modified simplex imbeddings method, which is related to the class of cutting plane methods. The main feature of this method is the convergence estimation, which depends only on the quantity of simplex vertices, that are cut off by the cutting plane. The more vertices are cut off by the cutting plane, the higher speed of method convergence. Modified method of simplex imbeddings with such criteria of cutting plane choosing is applied to solving special class of convex non-differentiable problems, which is consists of two types of functions. We come across the necessity to describe the function subdifferential that is depends on one or several parameters, that we can subject to optimization. It is described functions subdifferentials from the introduced class in parametric representation, that let us form auxiliary problems in simplex imbeddings method for searching resulting cutting planes, that cut off as much vertices of simplex as possible. It let us increase the speed of finding optimal solution. The results of numerical experiment are also given in this paper.
modified simplex imbeddings method, subdifferential of convex functions, resulting cutting plane.
PDF file (253 kB)
A. V. Kolosnitcyn, “Using of modified simplex imbeddings method for solving special class of convex non-differentiable optimization problems”, The Bulletin of Irkutsk State University. Series Mathematics, 11 (2015), 54–68
Citation in format AMSBIB
\paper Using of modified simplex imbeddings method for solving special class of~convex non-differentiable optimization problems
\jour The Bulletin of Irkutsk State University. Series Mathematics
Citing articles on Google Scholar:
Related articles on Google Scholar:
|Number of views:|