Covariate measurement error in generalized linear models (Q3756330)
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 4005359
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Covariate measurement error in generalized linear models |
scientific article; zbMATH DE number 4005359 |
Statements
Covariate measurement error in generalized linear models (English)
0 references
1987
0 references
errors-in-variables
0 references
EM algorithm
0 references
estimators of regression coefficients
0 references
generalized linear models
0 references
canonical link
0 references
normally distributed covariates
0 references
normally distributed measurement errors
0 references
missing data
0 references
iterative scheme
0 references
approximate first and second conditional moments
0 references
iteratively reweighted least squares
0 references
exact maximum likelihood solution
0 references
Gaussian quadrature
0 references
simple logistic regression with measurement error
0 references
0.8492966294288635
0 references
0.8369962573051453
0 references
0.8229827880859375
0 references
0.8191908001899719
0 references
The EM algorithm is used to obtain estimators of regression coefficients for generalized linear models with canonical link when normally distributed covariates are masked by normally distributed measurement errors. By casting the true covariates as 'missing data', the EM procedure suggests an iterative scheme in which each cycle consists of an E-step, requiring the computation of approximate first and second conditional moments of the true covariates given the observed data, followed by an M-step in which regression parameters are updated by iteratively reweighted least squares based on these approximations. The proposed procedure is compared numerically with the exact maximum likelihood solution, obtained by using Gaussian quadrature instead of the approximations in the E-step of the EM algorithm, and with alternative estimators for simple logistic regression with measurement error. The results for the proposed procedure are encouraging.
0 references