Events2Join

Controlled Markov chains with non|exponential discounting and ...


[2306.17448] Long-run impulse control with generalised discounting

... exponential) discounting on a long-run impulse control problem for a Feller-Markov process. We show that the optimal value of the discounted ...

Semi-Markov Control Models with Partially Known Holding Times ...

The paper deals with a class of semi-Markov control models with Borel state and control spaces and possibly unbounded costs.

Non-Stationary Markov Decision Processes, a Worst-Case ...

To define a Non-Stationary Markov Decision Process (NSMDP), we revert to the initial MDP model introduced by Puterman [2014], where the transition and reward ...

july 1986 lids-p-1583 control of markov chains with long-run ...

Existence of stable stationary strategies which are optimal in the appropriate sense is established and these are characterized via the dynamic programming.

Zero-sum continuous-time Markov games with unbounded transition ...

Our results are illustrated with a controlled queueing system with unbounded transition and reward rates. Keywords: controlled Q-process; discounted payoffs; ...

Estimation and Optimal Control for Constrained Markov Chains

Estimation and Optimal Control for Constrained Markov Chains by D.-j. Ma, A. Makowski, A. Shwartz. Full text available on Amanote Research.

Stability estimation of some Markov controlled processes - OUCI

Abstract We consider a discrete-time Markov controlled process endowed with the expected total discounted reward. We assume that the distribution of the ...

Controlled Markov Chains - World Scientific Publishing

Controlled Markov Chains · SOME LIMIT THEOREMS FOR COUNTABLE NON-HOMOGE-NEOUS MARKOV CHAINS · The Limit Properties of Relative Frequency of Markov Chains in ...

Influence and Variance of a Markov Chain : Application to Adaptive ...

We illustrate this method on the non-linear, two di- mensional “Car on the Hill” and the 4d “space-shuttle” and. “airplane-meeting” control problems. 1 ...

Markov chain - Wikipedia

A Markov chain or Markov process is a stochastic process describing a sequence of possible events in which the probability of each event depends only on the ...

[PDF] Continuous-Time Markov Chains | Semantic Scholar

A continuous-time Markov chain on the non ... This paper studies denumerable state continuous-time controlled Markov chains with the discounted reward criterion ...

Approximation of optimal ergodic dividend strategies using ...

By using the Markov chain approximation techniques, the authors construct a discrete-time controlled Markov chain for the approximation, and ...

Optimal Control of Probability on a Target Set for Continuous-Time ...

2012: Discounted Continuous-Time Controlled Markov Chains: Convergence of ... non-Markov decision processes International Journal of Systems Science 10 ...

A fuzzy approach to Markov decision processes with uncertain ...

The non-discounted reward problem for a controlled Markov set-chain was developed in [6,11]. This paper is organized as follows: In Section 2, ...

An Adaptive Markov Chain Approach for Probabilistic Forecasting of ...

a. First-order Markov Chain models · (t) denote the number of transitions between states of this single chain at time t − 1 and t, [i.e., Nij(t) = 1 when X(t − 1) ...

A Central Limit Theorem for Temporally Nonhomogenous Markov ...

The main theorem generalizes a classic result of Dobrushin for temporally nonhomogeneous Markov chains, and the principal innovation is that here the summands ...

Discrete-time control with non-constant discount factor - ProQuest

This paper deals with discrete-time Markov decision processes (MDPs) with Borel state and action spaces, and total expected discounted cost optimality ...

Markov Decision Processes and Dynamic Programming - Inria

For instance, in the control of an inverted pendulum, the state that can be observed is only the angular position θt. In this case the system is non-Markov ...

Technical Note—An Equivalence Between Continuous and Discrete ...

A continuous time Markov decision process with uniformly bounded transition rates is shown to be equivalent to a simpler discrete time Markov decision process.

Controlled Markov chains with constraints

Ross (1989) studied ergodic (or 'long-run average cost') control of Markov chains with finite state and action spaces when finitely many (say, m) other ergodic ...