scientific article; zbMATH DE number 3950345

From MaRDI portal
Publication:3720436





One of the most popular optimization problems for controlled Markov chains on an infinite time interval consists in the maximization of the limiting mean one-step payoff \[ W(u,X)=\lim_{T\to \infty}(1/T)\sum^{T}_{t=1}E_ uX_ t, \] where \(X_ t\) is the cost function, over the set of the admissible policies u. The author considers the generalization of the above problem assuming that u is such that W(u,Y)\(\leq g\) where Y is another cost function and g is a prescribed level. This problem is considered also in an adaptive setting.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3720436)