Weighted discounted Markov decision processes with perturbation

1.IntrodnctionTheweightedMarkovdecisionprocesses(MDP's)havebeenextensivelystudiedsince1980's,seeforinstance,[1-6]andsoon.ThetheoryofweightedMDP'swithperturbedtransitionprobabilitiesappearstohavebeenmentionedonlyin[7].Thispaperwilldiscussthemodelsofwe...     

Stochastic approximations of constrained discounted Markov decision processes

We consider a discrete-time constrained Markov decision process under the discounted cost optimality criterion. The state and action spaces are assumed to be Borel spaces, while the cost and constraint functions might be unbounded. We are interested in approximating numerically the optimal discounted constrained cost. To this end, we suppose that the transition kernel of the Markov decision process is absolutely continuous with respect to some probability measure μ  . Then, by solving the linear programming formulation of a constrained control problem related to the empirical probability measure _μn${\mu }_n$ of μ, we obtain the corresponding approximation of the optimal constrained cost. We derive a concentration inequality which gives bounds on the probability that the estimation error is larger than some given constant. This bound is shown to decrease exponentially in n. Our theoretical results are illustrated with a numerical application based on a stochastic version of the Beverton–Holt population model.     

A value iteration method for undiscounted multichain Markov decision processes

This paper proposes a value iteration method which finds an-optimal policy of an undiscounted multichain Markov decision process in a finite number of iterations. The undiscounted multichain Markov decision process is reduced to an aggregated Markov decision process, which utilizes maximal gains of undiscounted Markov decision sub-processes and is formulated as an optimal stopping problem. As a preliminary, sufficient conditions are presented under which a policy is-optimal.

---

Zusammenfassung In dieser Arbeit wird eine Wertiterationsmethode vorgeschlagen, die eine-optimale Politik für einen undiskontierten nicht-irreduziblen Markovschen Entscheidungsprozeß (MEP) in endlichen vielen Schritten liefert. Der undiskontierte nicht-irreduzible MEP wird auf einen aggregierten MEP reduziert, der maximale Gewinn eines undiskontierten Sub-MEP verwendet und als optimales Stopp-Problem formuliert wird. Zu Beginn werden hinreichende Bedingungen für die-Optimalität einer Politik angegeben.

     

A homotopy approach for infinite horizon discounted Markov decision processes

In this paper we consider a homotopy deformation approach to solving Markov decision process problems by the continuous deformation of a simpler Markov decision process problem until it is identical with the original problem. Algorithms and performance bounds are given.     

An envelope theorem and some applications to discounted Markov decision processes

In this paper, an Envelope Theorem (ET) will be established for optimization problems on Euclidean spaces. In general, the Envelope Theorems permit analyzing an optimization problem and giving the solution by means of differentiability techniques. The ET will be presented in two versions. One of them uses concavity assumptions, whereas the other one does not require such kind of assumptions. Thereafter, the ET established will be applied to the Markov Decision Processes (MDPs) on Euclidean spaces, discounted and with infinite horizon. As the first application, several examples (including some economic models) of discounted MDPs for which the et allows to determine the value iteration functions will be presented. This will permit to obtain the corresponding optimal value functions and the optimal policies. As the second application of the ET, it will be proved that under differentiability conditions in the transition law, in the reward function, and the noise of the system, the value function and the optimal policy of the problem are differentiable with respect to the state of the system. Besides, various examples to illustrate these differentiability conditions will be provided. This work was partially supported by Benemérita Universidad Aut ónoma de Puebla (BUAP) under grant VIEP-BUAP 38/EXC/06-G, by Consejo Nacional de Ciencia y Tecnología (CONACYT), and by Evaluation-orientation de la COopération Scientifique (ECOS) under grant CONACyT-ECOS M06-M01.     

A K-step look-ahead analysis of value iteration algorithms for Markov decision processes

We introduce and analyze a general look-ahead approach for Value Iteration Algorithms used in solving both discounted and undiscounted Markov decision processes. This approach, based on the value-oriented concept interwoven with multiple adaptive relaxation factors, leads to accelerating procedures which perform better than the separate use of either the concept of value oriented or of relaxation. Evaluation and computational considerations of this method are discussed, practical guidelines for implementation are suggested and the suitability of enhancing the method by incorporating Phase 0, Action Elimination procedures and Parallel Processing is indicated. The method was successfully applied to several real problems. We present some numerical results which support the superiority of the developed approach, particularly for undiscounted cases, over other Value Iteration variants.     

On some algorithms for limiting average Markov decision processes

We consider limiting average Markov decision processes (MDP) with finite state and action spaces. We propose some algorithms to determine optimal strategies for deterministic and general MDPs. These algorithms are based on graph theory and the construction of levels in some aggregated MDP.     

Average optimality for continuous-time Markov decision processes with a policy iteration approach

This paper deals with the average expected reward criterion for continuous-time Markov decision processes in general state and action spaces. The transition rates of underlying continuous-time jump Markov processes are allowed to be unbounded, and the reward rates may have neither upper nor lower bounds. We give conditions on the system's primitive data and under which we prove the existence of the average reward optimality equation and an average optimal stationary policy. Also, under our conditions we ensure the existence of ?-average optimal stationary policies. Moreover, we study some properties of average optimal stationary policies. We not only establish another average optimality equation on an average optimal stationary policy, but also present an interesting “martingale characterization” of such a policy. The approach provided in this paper is based on the policy iteration algorithm. It should be noted that our way is rather different from both the usually “vanishing discounting factor approach” and the “optimality inequality approach” widely used in the previous literature.     

Modified policy iteration algorithms are not strongly polynomial for discounted dynamic programming

This note shows that the number of arithmetic operations required by any member of a broad class of optimistic policy iteration algorithms to solve a deterministic discounted dynamic programming problem with three states and four actions may grow arbitrarily. Therefore any such algorithm is not strongly polynomial. In particular, the modified policy iteration and $\lambda$-policy iteration algorithms are not strongly polynomial.     

Hierarchical algorithms for discounted and weighted Markov decision processes

We consider a discrete time finite Markov decision process (MDP) with the discounted and weighted reward optimality criteria. In [1] the authors considered some decomposition of limiting average MDPs. In this paper, we use an analogous approach for discounted and weighted MDPs. Then, we construct some hierarchical decomposition algorithms for both discounted and weighted MDPs.     

The value iteration algorithm is not strongly polynomial for discounted dynamic programming

This note provides a simple example demonstrating that, if exact computations are allowed, the number of iterations required for the value iteration algorithm to find an optimal policy for discounted dynamic programming problems may grow arbitrarily quickly with the size of the problem. In particular, the number of iterations can be exponential in the number of actions. Thus, unlike policy iterations, the value iteration algorithm is not strongly polynomial for discounted dynamic programming.     

Monotonicity in multidimensional Markov decision processes for the batch dispatch problem

Structural properties of stochastic dynamic programs are essential to understanding the nature of the solutions and in deriving appropriate approximation techniques. We concentrate on a class of multidimensional Markov decision processes and derive sufficient conditions for the monotonicity of the value functions. We illustrate our result in the case of the multiproduct batch dispatch (MBD) problem.     

Adaptive control of discounted Markov decision chains

In this paper, we consider discounted-reward finite-state Markov decision processes which depend on unknown parameters. An adaptive policy inspired by the nonstationary value iteration scheme of Federgruen and Schweitzer (Ref. 1) is proposed. This policy is briefly compared with the principle of estimation and control recently obtained by Schäl (Ref. 4).This research was supported in part by the Consejo Nacional de Ciencia y Tecnología under Grant No. PCCBBNA-005008, in part by a grant from the IBM Corporation, in part by the Air Force Office of Scientific Research under Grant No. AFOSR-79-0025, in part by the National Science Foundation under Grant No. ECS-0822033, and in part by the Joint Services Electronics Program under Contract No. F49620-77-C-0101.     

Accelerated modified policy iteration algorithms for Markov decision processes

We propose a new approach to accelerate the convergence of the modified policy iteration method for Markov decision processes with the total expected discounted reward. In the new policy iteration an additional operator is applied to the iterate generated by Markov operator, resulting in a bigger improvement in each iteration.     

Block-successive approximation for a discounted Markov decision model

In this paper we suggest a new successive approximation method to compute the optimal discounted reward for finite state and action, discrete time, discounted Markov decision chains. The method is based on a block partitioning of the (stochastic) matrices corresponding to the stationary policies. The method is particularly attractive when the transition matrices are jointly nearly decomposable or nearly completely decomposable.