Prediction of Random Effects in Linear and Generalized Linear Models under Model Misspecification
β Scribed by Charles E. McCulloch; John M. Neuhaus
- Book ID
- 109224374
- Publisher
- John Wiley and Sons
- Year
- 2010
- Tongue
- English
- Weight
- 358 KB
- Volume
- 67
- Category
- Article
- ISSN
- 0006-341X
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
The meta-analysis of multi-centre trials can be based on either "xed or random e!ect models. This paper argues for the general use of random e!ect models, and illustrates the value of non-parametric maximum likelihood (NPML) analysis of such trials. The same general approach uni"es administrative &l
Assuming a general linear model with known covariance matrix, several linear and nonlinear predictors are presented and their properties are discussed. In the context of simultaneous multiple prediction, a total sum of squared errors is suggested as a loss function for comparing predictors. Based on