|
Journal of Convex Analysis 11 (2004), No. 1, 163--178 Copyright Heldermann Verlag 2004 Closing the Duality Gap in Linear Vector Optimization Andreas H. Hamel Martin-Luther-University Halle-Wittenberg, Dept. of Mathematics and Computer Science, Theodor-Lieser-Str. 5, 06099 Halle, Germany, hamel@mathematik.uni-halle.de Frank Heyde Martin-Luther-University Halle-Wittenberg, Dept. of Mathematics and Computer Science, Theodor-Lieser-Str. 5, 06099 Halle, Germany Andreas Löhne Martin-Luther-University Halle-Wittenberg, Dept. of Mathematics and Computer Science, Theodor-Lieser-Str. 5, 06099 Halle, Germany Christiane Tammer Martin-Luther-University Halle-Wittenberg, Dept. of Mathematics and Computer Science, Theodor-Lieser-Str. 5, 06099 Halle, Germany Kristin Winkler Martin-Luther-University Halle-Wittenberg, Dept. of Mathematics and Computer Science, Theodor-Lieser-Str. 5, 06099 Halle, Germany Using a set-valued dual cost function we give a new approach to duality theory for linear vector optimization problems. We develop the theory very close to the scalar case. Especially, in contrast to known results, we avoid the appearance of a duality gap in case of b = 0 . Examples are given. Keywords: set-valued optimization, duality, linear multicriteria optimization. MSC: 90C29, 90C46, 90C05. FullText-pdf (362 KB) for subscribers only. |