|
|
|
Vietnam Journal of Mathematics 35:4(2007)
333-414
|
|
Comparison of Search Strategies of Branch and
Bound Algorithm for Concave Minimization Problems Under Linear Constraints
|
|
Hiroshi Konno,
Kazuo Izumi, and Rei Yamamoto
|
|
Abstract. Solving large scale concave
minimization problems is attracting more attention in recent years. A
number of large scale practical problems have been solved to optimality by
a branch and bound algorithm based on hyper-rectangular subdivision/linear
under-estimating function and by a 0-1 mixed linear integer programming
algorithm.We will report the result of a systematic comparison of search
strategies of branch and bound algorithm. Also,we compare them with 0-1
integer programming approach.We will show that a depth first search
strategy is useful for solving a very large scale mean-absolute deviation
portfolio optimization problem under concave transaction cost.
|
|
|
|
Keywords: Global optimization, concave
minimization, branch and bound algorithm, mixed 0-1 programming, portfolio
optimization, mean-absolute deviation model.
|
|
|
|
Established
by Vietnam Academy of Science and Technology & Vietnam Mathematical
Society
Published
by Springer since January 2013
|
|
