No label defined (Q4730638)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 4115776
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 4115776 |
Statements
1989
0 references
finite-state Markov chain
0 references
vector-valued process
0 references
second-order moments
0 references
spectral density
0 references
stationary probabilities
0 references
mean
0 references
maximum likelihood estimator
0 references
asymptotically efficient
0 references
numerical-valued Markov chain
0 references
wide- sense first-order autoregressive process
0 references
0.9012219
0 references
0.89730716
0 references
0.8945194
0 references
0.88947487
0 references
0.8804336
0 references
0.87409264
0 references
0.87251246
0 references
[For the entire collection see Zbl 0679.00013.] A finite-state Markov chain is a stochastic process in which the variable takes on one of a finite number of values. Although the values may be numerical, they need not be; they may be simply states or categories. If they are given numerical values, the values are not necessarily the first so many integers, as in the number of customers waiting in a queue. When the chain is described in terms of states, it is convenient for many purposes to treat the chain as a vector-valued process. The vector has 1 in the position corresponding to the given state and 0 in the other positions. Then the vector-valued process is first-order autoregressive in the wide sense when the Markov chain is first-order. In this paper, the consequences of the autoregressive structure of the vector-valued process are developed further to yield various second-order moments and the spectral density of the process. It is shown that to estimate the stationary probabilities from a finite sequence of observations the mean of the these observations is asymptotically equivalent to the maximum likelihood estimator and is asymptotically efficient. The numerical-valued Markov chain is considered as a linear function of the vector-valued process, and a simple condition is obtained for it to be a wide-sense first-order autoregressive process.
0 references