Abstract

The paper presents a finite branch-and-bound variant of an outcome-based algorithm proposed by \textit{H. P. Benson} and \textit{D. Lee} [J. Optimization Theory Appl. 88, 77-105 (1996; Zbl 0842.90099)] for minimizing a lower-semicontinuous function over the efficient set of a bicriteria linear programming problem. Similarly to the Benson-Lee algorithm, we work primarily in the outcome space. Dissimilarly, instead of constructing a sequence of consecutive efficient edges in the outcome space, we use the idea of generating a refining sequence of partitions covering the at most two-dimensional efficient set in the outcome space. Computational experience is also presented.