On ordered stopping times of a Markov process (Q2641010): Difference between revisions

Let X be a transient strong Markov process with potential kernel U. It is well-known that if \(\nu\) \(U\leq \mu U\), then there exists a stopping time T such that \(\nu (\cdot)=P^{\mu}(X_ T\in \cdot)\). The author extends this result to the case of infinite sequences and shows that if \(\nu_ 1U\leq \nu_ 2U\leq...\leq \mu U\), then there exists a decreasing sequence of randomized stopping times \(T_ n\) such that \(\nu_ n\) is the law of \(X_{T_ n}\) under the initial distribution \(\mu\). For the proof, he uses the Baxter-Chacon compactness theorem for randomized stopping times. The problem considered in this paper has also been solved by \textit{C. T. Shih} [Ann. Probab. 18, No.4, 1623-1634 (1990)] by a different method.