APPROXIMATE SOLUTIONS OF INTERVAL-VALUED OPTIMIZATION PROBLEMS

Abstract

This paper deals with approximate solutions of an optimization problem with interval-valued ob- jective function. Four types of approximate solution concepts of the problem are proposed by con- sidering the partial ordering LU on the set of all closed and bounded intervals. We show that these solutions exist under very weak conditions. Under suitable constraint qualifications, we de-
rive Karush–Kuhn–Tucker necessary and sufficient optimality conditions for convex interval-valued optimization problems.