The following notes are largely written by me, so I take responsibility for any errors. Some of the code is based on material from previous iterations of the course.
Notes for discussion section
- Lab 1 (September 1)
- Lab 2 (September 8)
- Lab 3 (September 15)
- Lab 4 (September 22)
- Lab 5 (September 29)
- One-way ANOVA and sequential ANOVA [HTML]
- Discuss review questions for Midterm 1
- Lab 6 (October 5)
- Leverage, Mahalanobis distance, added variable plots, component-plus-residual plots [HTML]
- Lab 7 (October 12)
- Regression diagnostics: outliers, residuals (standardized, predicted, standardized predicted), Cook’s distance [HTML]
- Lab 8 (October 19)
- Regression diagnostics: variance-stabilizing transformations, Q-Q plots [HTML]
- Lab 9 (October 26)
- Variable selection (forward/backward stepwise, adjusted $R^2$, AIC, BIC, Mallows’s $C_p$) [HTML]
- Ant dataset discussion
- Lab 10 (November 3)
- Lab 11 (November 17)
- Generalized linear models (review, residual deviance, null deviance,
glm
, prediction example with logistic model, choosing a threshold, ROC curve) [HTML]
- Generalized linear models (review, residual deviance, null deviance,
- Lab 12 (December 1)
- Classification trees (impurity, reading
rpart
output,printcp
) and shrinkage (ridge regression, LASSO, elastic net,glmnet
) [HTML]
- Classification trees (impurity, reading