A Method for Minimization of Quasidifferentiable Functions (Q4806341)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1910440
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | A Method for Minimization of Quasidifferentiable Functions |
scientific article; zbMATH DE number 1910440 |
Statements
A Method for Minimization of Quasidifferentiable Functions (English)
0 references
2002
0 references
subdifferential
0 references
quasidifferential
0 references
D.C functions
0 references
0 references
0 references
0 references
0.858491063117981
0 references
0.8454769849777222
0 references
0.8353965282440186
0 references
0.8352864980697632
0 references
Let \(f\) be a nonsmooth function which is presented as a difference of two convex functions \(f_1\) and \(f_2\) on \(\mathbb{R}^n.\) It is known from the quasidifferential theory of Demyanov and Rubinov that every minimum point of \(f\) on \(\mathbb{R}^n\) is stationary in the sense that \(\partial f_2(x)\subseteq \partial f_1(x)\) where \(\partial f_i(x)\) is the subdifferential of \(f_i\) at \(x, i=1,2\). In this paper the author presents a method to find stationary points of \(f\). The main step of the method is the computation of descent directions. Two algorithms are proposed to perform this step. They are based on continuous approximations of the quasidifferential of \(f\) which can be calculated by discrete gradients of the function. The convergence of the method is proven and some numerical experiments are given.
0 references