A NONLINEAR-PROGRAMMING MODEL FOR PARTIALLY OBSERVABLE MARKOV DECISION-PROCESSES - FINITE-HORIZON CASE

1995-11-02
The concept of partially observable Markov decision processes was born to handle the problem of lack of information about the state of a Markov decision process. If the state of the system is unknown to the decision maker then an obvious approach is to gather information that is helpful in selecting an action, This problem was already solved using the theory of Markov processes. We construct a nonlinear programming model for the same problem and develop a solution algorithm that turns out to be a policy iteration algorithm. The policies found this way are easier to use than the policies found by the existing method, although they have the same optimal objective value.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH

Suggestions

A Stochastic Maximum Principle for a Markov Regime-Switching Jump-Diffusion Model with Delay and an Application to Finance
Savku, Emel; Weber, Gerhard Wilhelm (2018-11-01)
We study a stochastic optimal control problem for a delayed Markov regime-switching jump-diffusion model. We establish necessary and sufficient maximum principles under full and partial information for such a system. We prove the existence-uniqueness theorem for the adjoint equations, which are represented by an anticipated backward stochastic differential equation with jumps and regimes. We illustrate our results by a problem of optimal consumption problem from a cash flow with delay and regimes.
Employing decomposable partially observable Markov decision processes to control gene regulatory networks
Erdogdu, Utku; Polat, Faruk; Alhajj, Reda (2017-11-01)
Objective: Formulate the induction and control of gene regulatory networks (GRNs) from gene expression data using Partially Observable Markov Decision Processes (POMDPs).
An Interactive partitioning approach for multiobjective decision making under a general monotone utility function
Karasakal, Esra (2013-09-01)
We develop an interactive partitioning approach for solving the multiobjective decision making problem of a decision maker (DM) who has an implicit general monotone utility function. The approach reduces feasible solution space using the DM's preferences. Hypothetical solutions called partition ideals (PIs) that dominate portions of the efficient frontier are generated and those that are inferior to a feasible solution are used to eliminate the dominated regions. We investigate the issues in representation ...
A SPINOR MODEL FOR QUANTUM COSMOLOGY
DERELI, T; ONDER, M; TUCKER, RW (1994-03-31)
The question of the interpretation of Wheeler-DeWitt solutions in the context of cosmological models is addressed by implementing the Hamiltonian constraint as a spinor wave equation in minisuperspace. We offer a relative probability interpretation based on a non-closed vector current in this space and a prescription for a parametrisation of classical solutions in terms of classical time. Such a prescription can accommodate classically degenerate metrics describing manifolds with signature change. The relat...
A two-level variational multiscale method for convection-dominated convection-diffusion equations
Volker, John; Kaya Merdan, Songül; Layton, William (2006-01-01)
This paper studies the error in, the efficient implementation of and time stepping methods for a variational multiscale method (VMS) for solving convection-dominated problems. The VMS studied uses a fine mesh C-O finite element space X-h to approximate the concentration and a coarse mesh discontinuous vector finite element space L-H for the large scales of the flux in the two scale discretization. Our tests show that these choices lead to an efficient VMS whose complexity is further reduced if a (locally) L...
Citation Formats
Y. Y. Serin, “A NONLINEAR-PROGRAMMING MODEL FOR PARTIALLY OBSERVABLE MARKOV DECISION-PROCESSES - FINITE-HORIZON CASE,” EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, pp. 549–564, 1995, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/47690.