An envelope theorem and some applications to discounted Markov decision processes (Q2483011): Difference between revisions

This paper considers an Envelope Theorem for optimization problems on the Euclidean space under two different class conditions: One is a concavity assumptions, the other are differentiability conditions in the transition law, in the reward function, and the noise of the system. The criterion in this paper is expected total discounted reward. Some interesting examples of economic models are also presented.