Aggregation of Conditional Absorbing Markov Chains.
Abstract
When modeling a process by means of a finite Markov chain, it is sometimes necessary or desirable to stratify the process into subprocesses and model each of these individual subprocesses. For example, in a study of a distributed data base system (1) the flow of data was modeled as a Markov chain for several separate geographic locations. In a recruiting study (2) the movement of military-age men through the recruiting process and into the armed forces was modeled for separate racial and educational groups as a Markov chain with a single state space for each group--only the input data (transition probabilities) to the model were changed for each group. In (3) a Markov chain model was used to investigate the consequences of induced abortion for different groups of women by estimating transition probabilities separately for each group.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1982
- Accession Number
- ADA116603
Entities
People
- C. Bernard Barfoot
Organizations
- Center for Naval Analyses