Isye 6414 Final Study Guide Rated A+

4. The sampling distribution of the prediction of future responses is a t-distribution under the Poisson regression model. {Ans: False}18. Cross-validation using random sampling is both more computationally efficient and more accurate in estimating the classification error than the K-fold cross validation. {Ans: False}6. We cannot perform a residual analysis for Poisson regression. {Ans: False}2. Penalization in linear regression models means penalizing for complex models, that is, models with a large number of predictors. {Ans: True}14. The error term in logistic regression has a normal distribution. {Ans: False}8. The lasso regression requires a numerical algorithm to minimize the penalized sum of least squares. {Ans: True}