SOME THEOREMS ON FUNCTIONALS OF MARKOV CHAINS,
Abstract
Various phenomena associated with a Markov process in discrete time are investigated. Attention is focused on recurrent events (i.e., to successive entrances into some fixed state of a Markov chain with the positive integers as states). Working time distribution is completely determined by the sequence (EY sub n), where Y sub n is the time, as observed from n, that the event last took place. Criteria for the event to be persistnet, transient, positive, etc., may be given directly in terms of the EY sub n. A particular class of null events called betaregular is examined and various joint-limit distributions are found for some of the functionals usually associated with these events. These limit laws are extended to situations more general than recurrent events, and these extended results are then applied to several concrete situations. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1964
- Accession Number
- AD0438439
Entities
People
- Sidney C. Port
Organizations
- RAND Corporation