Quasi-concave minimization subject to linear constraints

AbstractA finite algorithm is presented for solving the quasi-concave minimization problem subject to linear constraints. The concept of an extreme point is generalized to that of an extreme facet of a polyhedron. Then a search routine is developed for the detection of an extreme facet of the feasible region relative to the polyhedron defined by the current set of cuts. After identifying an extreme facet we cut it off by a cut developed for this purpose. We call this cut the facet cut. The method is both compatible with other cutting procedures and is finite.