Restricted maximum likelihood: Difference between revisions
simplify start up |
m space |
||
| Line 1: | Line 1: | ||
In [[statistics]], the '''restricted''' (or '''residual''', or '''reduced''') '''maximum likelihood''' ('''REML''') approach is a particular form of [[maximum likelihood]] estimation which does not base estimates on a maximum likelihood fit of all the information, but instead uses a [[likelihood function]] calculated from a transformed set of data, so that[[nuisance parameter]]s have no effect.<ref name=Dodge>Dodge, Y. (2003) ''The Oxford Dictionary of Statistical Terms'', OUP, ISBN 0199206139 (see REML)</ref> |
In [[statistics]], the '''restricted''' (or '''residual''', or '''reduced''') '''maximum likelihood''' ('''REML''') approach is a particular form of [[maximum likelihood]] estimation which does not base estimates on a maximum likelihood fit of all the information, but instead uses a [[likelihood function]] calculated from a transformed set of data, so that [[nuisance parameter]]s have no effect.<ref name=Dodge>Dodge, Y. (2003) ''The Oxford Dictionary of Statistical Terms'', OUP, ISBN 0199206139 (see REML)</ref> |
||
In the case of [[variance component]] estimation, the original data set is replaced by a set of [[contrast (statistics)|contrasts]] calculated from the data, and the likelihood function is calculated from the probability distribution of these contrasts, according to the model for the complete data set. In particular, REML is used as a method for fitting linear [[mixed model]]s. In contrast to the earlier [[maximum likelihood]] estimation, REML can produce [[unbiased]] estimates of variance and covariance parameters.<ref> |
In the case of [[variance component]] estimation, the original data set is replaced by a set of [[contrast (statistics)|contrasts]] calculated from the data, and the likelihood function is calculated from the probability distribution of these contrasts, according to the model for the complete data set. In particular, REML is used as a method for fitting linear [[mixed model]]s. In contrast to the earlier [[maximum likelihood]] estimation, REML can produce [[unbiased]] estimates of variance and covariance parameters.<ref> |
||
Revision as of 17:30, 13 August 2010
In statistics, the restricted (or residual, or reduced) maximum likelihood (REML) approach is a particular form of maximum likelihood estimation which does not base estimates on a maximum likelihood fit of all the information, but instead uses a likelihood function calculated from a transformed set of data, so that nuisance parameters have no effect.[1]
In the case of variance component estimation, the original data set is replaced by a set of contrasts calculated from the data, and the likelihood function is calculated from the probability distribution of these contrasts, according to the model for the complete data set. In particular, REML is used as a method for fitting linear mixed models. In contrast to the earlier maximum likelihood estimation, REML can produce unbiased estimates of variance and covariance parameters.[2]
The idea underlying REML estimation was put forward by M. S. Bartlett in 1937.[1][3] The first description of the approach applied to estimating components of variance in unbalanced data was by Desmond Patterson and Robin Thompson[1][4] of the University of Edinburgh, although they did not use the term REML. A review of the early literature was given by Harville.[5]
REML estimation is available in a number of general-purpose statistical software packages, including Genstat (the REML directive), SAS (the MIXED procedure), SPSS (the MIXED command), Stata (the xtmixed command), and R (the lme4 and older nlme packages) , as well as in more specialist packages such as MLwiN, HLM and ASReml.
References
- ^ a b c Dodge, Y. (2003) The Oxford Dictionary of Statistical Terms, OUP, ISBN 0199206139 (see REML)
- ^ Baker, Bob. Estimating variances and covariances
- ^ Bartlett, M.S. (1937). "Properties of sufficiency and statistical tests". Proceedings of the Royal Society of London, Series A. 160: 268–282. doi:10.1098/rspa.1937.0109.
- ^ Patterson, H.D.; Thompson, R. (1971). "Recovery of inter-block information when block sizes are unequal". Biometrika. 58 (3): 545–554. doi:10.1093/biomet/58.3.545. MR0319325.
- ^ Harville, David A. (1977). "Maximum Likelihood Approaches to Variance Component Estimation and to Related Problems". Journal of the American Statistical Association. 72 (358). Journal of the American Statistical Association, Vol. 72, No. 358: 320–338. doi:10.2307/2286796. MR0451550.
 | This statistics-related article is a stub. You can help Wikipedia by expanding it. |