Please use this identifier to cite or link to this item:
Fischer, Markus
Reiß, Markus
Year of Publication: 
Series/Report no.: 
SFB 649 discussion paper 2005,038
As a main step in the numerical solution of control problems in continous time, the controlled process is approximated by sequences of controlled Markov chains, thus discretizing time and space. A new feature in this context is to allow for delay in the dynamics. The existence of an optimal strategy with respect to the cost functional can be guaranteed in the class of relaxed controls. Weak convergence of the approximating extended Markov chains to the original process together with convergence of the associated optimal strategies is established.
Document Type: 
Working Paper

Files in This Item:
522.13 kB

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.