Efficient Screening of Nonnormal Regression Models

摘要: We consider some nonnormal regression situations in which there are many regressor variables, and it is desired to determine good fitting models, according the value of likelihood ratio statistic for tests submodels against full model. EfXicient computational algorithms normal linear model adopted use with models. Even as 10-15 variables present, wefind often possible all betterfitting models relatively small amounts computer time. The procedures illustrated on exponential, Poisson binary