A finite method for globally minimizing concave functions over unbounded polyhedral convex sets and its applications (Q1079499)

The problem considered is clearly described in the title. The function is \(f: {\mathbb{R}}^ n\to {\mathbb{R}}\) and the constraint set is of the form \(D=\{x:\) Ax\(\leq b\), \(x\geq 0\}\). For the case of a bounded D a finite algorithm has been developed by the author, \textit{Bui The Tam} and \textit{Vu Thien Ban} [ibid. 8, No.1, 21-40 (1983; Zbl 0562.90069)] which is suitable for applications in which ''we have to treat a sequence of linearly constrained concave minimization problems each of which differs from the previous one by one additional constraint'' (for instance, decomposition techniques). This paper, which is self-contained, extends those results (preserving finiteness) to the case when D is not necessarily bounded. Detailed background and references of previous work by the author and others are provided as well as a small numerical example and a summary of some computational experience.