This paper deals with a special case of Bayesian estimation in Markov processes. We consider a two state process where one transition probability is known exactly, while the other is assumed Beta distributed. Under these conditions the expected values of the steady state probabilities are obtained through the use of the hypergeometric function, a mathematical function heretofore encountered only in an entirely different area of applied mathe¬matics. Knowing the expected values of the steady state probabilities enables us to place the process considered into a statistical decision framework.
展开▼