Star-free language: Difference between revisions
en>Helpful Pixie Bot m ISBNs (Build KC) |
Deltahedron (talk | contribs) cite Lawson (2004) |
||
| Line 1: | Line 1: | ||
In [[statistical theory]], a '''pseudolikelihood''' is an [[approximation]] to the [[joint probability distribution]] of a collection of [[random variable]]s. The practical use of this is that it can provide an approximation to the [[likelihood function]] of a set of observed data which may either provide a computationally simpler problem for [[Estimation theory|estimation]], or may provide a way of obtaining explicit estimates of model parameters. | |||
The pseudolikelihood approach was introduced by [[Julian Besag]]<ref>Besag, J. (1975) "Statistical Analysis of Non-Lattice Data." ''The Statistician'', 24(3), 179–195</ref> in the context of analysing data having [[spatial dependence]]. | |||
==Definition== | |||
Given a set of random variables <math>X = X_1, X_2, ... X_n</math> and a set <math>E</math> of dependencies between these random variables, where <math> \lbrace X_i,X_j \rbrace \notin E </math> implies <math>X_i</math> is [[Conditional independence|conditionally independent]] of <math>X_j</math> given <math>X_i</math>'s neighbors, the pseudolikelihood of <math>X = x = (x_1,x_2, ... x_n)</math> is | |||
:<math>\Pr(X = x) = \prod_i \Pr(X_i = x_i|X_j = x_j\ \mathrm{for\ all\ } j\ \mathrm{for\ which}\ \lbrace X_i,X_j \rbrace \in E).</math> | |||
Here <math>X</math> is a vector of variables, <math>x</math> is a vector of values. The expression <math>X = x</math> above means that each variable <math>X_i</math> in the vector <math>X</math> has a corresponding value <math>x_i</math> in the vector <math>x</math>. The expression <math>\Pr(X = x)</math> is the probability that the vector of variables <math>X</math> has values equal to the vector <math>x</math>. Because situations can often be described using state variables ranging over a set of possible values, the expression <math>\Pr(X = x)</math> can therefore represent the probability of a certain state among all possible states allowed by the state variables. | |||
The '''pseudo-log-likelihood''' is a similar measure derived from the above expression. Thus | |||
:<math>\log \Pr(X = x) = \sum_i \log \Pr(X_i = x_i|X_j = x_j\ \mathrm{for\ all}\ \lbrace X_i,X_j \rbrace \in E).</math> | |||
One use of the pseudolikelihood measure is as an approximation for inference about a [[Markov random field|Markov]] or [[Bayesian network]], as the pseudolikelihood of an assignment to <math>X_i</math> may often be computed more efficiently than the likelihood, particularly when the latter may require marginalization over a large number of variables. | |||
==Properties== | |||
Use of the pseudolikelihood in place of the true likelihood function in a [[maximum likelihood]] analysis can lead to good estimates, but a straightforward application of the usual likelihood techiques to derive information about estimation uncertainty, or for [[Statistical significance|significance testing]], would in general be incorrect.<ref>Dodge, Y. (2003) ''The Oxford Dictionary of Statistical Terms'', OUP. ISBN 0-19-920613-9</ref> | |||
== References == | |||
{{reflist}} | |||
[[Category:Statistical inference]] | |||
Revision as of 21:26, 27 November 2012
In statistical theory, a pseudolikelihood is an approximation to the joint probability distribution of a collection of random variables. The practical use of this is that it can provide an approximation to the likelihood function of a set of observed data which may either provide a computationally simpler problem for estimation, or may provide a way of obtaining explicit estimates of model parameters.
The pseudolikelihood approach was introduced by Julian Besag[1] in the context of analysing data having spatial dependence.
Definition
Given a set of random variables and a set of dependencies between these random variables, where implies is conditionally independent of given 's neighbors, the pseudolikelihood of is
Here is a vector of variables, is a vector of values. The expression above means that each variable in the vector has a corresponding value in the vector . The expression is the probability that the vector of variables has values equal to the vector . Because situations can often be described using state variables ranging over a set of possible values, the expression can therefore represent the probability of a certain state among all possible states allowed by the state variables.
The pseudo-log-likelihood is a similar measure derived from the above expression. Thus
One use of the pseudolikelihood measure is as an approximation for inference about a Markov or Bayesian network, as the pseudolikelihood of an assignment to may often be computed more efficiently than the likelihood, particularly when the latter may require marginalization over a large number of variables.
Properties
Use of the pseudolikelihood in place of the true likelihood function in a maximum likelihood analysis can lead to good estimates, but a straightforward application of the usual likelihood techiques to derive information about estimation uncertainty, or for significance testing, would in general be incorrect.[2]
References
43 year old Petroleum Engineer Harry from Deep River, usually spends time with hobbies and interests like renting movies, property developers in singapore new condominium and vehicle racing. Constantly enjoys going to destinations like Camino Real de Tierra Adentro.