摘要：In this paper, we develop two algorithms for globally optimizing a special
class of linear programs with an additional concave constraint. We assume that the
concave constraint is defined by a separable concave function. Exploiting this special
structure, we apply Falk-Soland's branch-and-bound algorithm for concave minimization
in both direct and indirect manners. In the direct application, we solve the problem
alternating local search and branch-and-bound. In the indirect application, we carry
out the bounding operation using a parametric right-hand-side simplex algorithm.
