Discussion work (30 mins):

Looking at the BloodBrain data in the caret package, perform variable selection on this dataset, using both LASSO (you can use the glmnet package for this) and step AIC (which is implemented in the MASS package). Compare the models selected by the different methods.

Recall that the BloodBrain data consists of a vector logBBB of response values and a matrix bbbDescr of predictors.