Maxentropy completion and properties of some partially defined Stationary Markov chains

Abstract

We consider a stationary Markovian evolution with values on a disjointly partitioned set space $I\bigsqcup \mathcal{E}$. The evolution is visible (in the sense of knowing the transition probabilities) on the states in $I$ but not for the states in $\mathcal{E}$. One only knows some partial information on the transition probabilities on $\mathcal{E}$, the input and output transition probabilities and some constraints of the transition probabilities on $\mathcal{E}$. Under some conditions we supply the transition probabilities on $\mathcal{E}$ that satisfies the maximum entropy principle.

Section 4 modified, misprints corrected