|
|
|
Vietnam Journal of Mathematics 35:4(2007) 463-479
|
|
A Simplicial Algorithm for Concave
Minimization and Its Performance
as a Heuristic Tool
|
|
Takahito Kuno and Yoshiyuki Shiguro
|
|
Abstract.
In this paper, we develop a kind of branch-and-bound algorithm
for solving concave minimization problems.
We show that the algorithm converges to an optimal solution
of this multiextremal global optimization problem,
and that it generates a high-quality heuristic
solution even if it is forced to terminate.
Therefore, the algorithm can be used in two ways,
as an exact algorithm and as a heuristic tool.
We also report some numerical results of a comparison
with an existing algorithm,
and show the performance as a heuristic tool.
|
|
|
|
|
|
Keywords: Global optimization, concave minimization,
branch-and-bound algorithm, simplicial algorithm,
heuristic algorithm.
|
|