A scaled stochastic approximation algorithm
Management Science
Comparative study of stochastic algorithms for system optimization based on gradient approximations
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Hi-index | 0.00 |
In this document we propose a discrete time Markov decision process with finite state to represent some stochastic and dynamical systems. Our problem consists on finding the optimal policy that maximizes the expected average reward per unit of time under an infinite planning horizon using stochastic linear programming. We analyze the feasibility and optimality properties of the model allowing that some of the elements of the A matrix of technological coefficients to be random. Our aim is to enable the transition probability matrix thinking of substituting it punctual values by some probability density functions. We report the theoretical results and some numeric examples.