(g, w)—Optima in Markov Renewal Programs

Abstract

Policies nearly optimal for small discount rates (more precisely, ( g, w )—optimal policies) are found for completely ergodic Markov Renewal Programs by linear programming.