Strategies for Model Mixing in Generalized Linear Models
Proceedings of the Sixth International Workshop on Artificial Intelligence and Statistics, PMLR R1:103-114, 1997.
In linear regression models and generalized linear regression models (GLMs), there is often substantial uncertainty about the choice of covariates to include in the model. Both classical and Bayesian approaches that involve selecting a subset of covariates and making inferences conditional on that model choice ignore a major component of uncertainty in the problem. One approach for incorporating this form of model uncertainty into the analysis is by directly building into the model a vector of indicator variables $Y$ that reflects which covariates are included in the model ...