We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
On the Modification of the Discrete Filled Function Algorithm for Nonlinear Discrete Optimization.
- Authors
Woon, S. F.; Karim, S.; Mohamad, M. S. A.; Ryan, L.; Rehbock, V.
- Abstract
The discrete filled function method (DFFM) is a global optimization method for searching for the best solution amongst multiple local optima. This method consists of two phases: in the first phase, an ordinary descent method is used to find a local minimum; in the second phase, an auxiliary function, called a filled function, is introduced that has a maximizer at the current local minimum, so that minimizing the filled function leads to improved points. Once an improved point is found, it can serve as a starting point for the next local search. In this paper, we consider a standard discrete filled function algorithm in the literature and propose a modification to increase its efficiency. Three numerical examples are given to demonstrate the proposed modification's potential in solving large scale discrete optimization problems.
- Subjects
NONLINEAR functions; GLOBAL optimization; ALGORITHMS; MAXIMA &; minima
- Publication
IAENG International Journal of Applied Mathematics, 2021, Vol 51, Issue 4, p930
- ISSN
1992-9978
- Publication type
Article