Infinite horizon programs; convergence of approximate solutions |
| |
Authors: | Sjur D Flåm Alain Fougères |
| |
Institution: | (1) Economics Department, University of Bergen, 5008, Norway;(2) Mathematics Department, University of Perpignan, 66025 Cedex, France |
| |
Abstract: | This paper deals with infinite horizon, dynamic programs, stated in discrete time, and afflicted by no uncertainty. The essential objective, to be minimized, is the accumulated value of all discounted future costs, and it is assumed to satisfy the crucial condition that every lower level set is bounded with respect to a certain norm. That norm, as well as the natural space of trajectories, is problem intrinsic.In contrast to standard Markov decision processes (MDP) we admit unbounded singleperiod cost functions and exponential growth within an unlimited state space. Also, no assumption about stationarity in problem data is made.We show, under broad hypotheses, that any minimizing sequence accumulates to points which solve the dynamic program optimally. This result is important for the study of approximation schemes.Supported by grants from Total Marine via NTNF, and Wilhelm Keilhau's Minnefond. |
| |
Keywords: | Growth conditions convex analysis calculus of variations Bolza problems Markov decision processes dynamic programming |
本文献已被 SpringerLink 等数据库收录! |
|