#forward-selection

Question 1: Select all that apply

You compare feature selection methods: forward selection, backward selection, recursive feature elimination (RFE) and exhaustive search. Which statements are true? Select all that apply.

Forward selection is guaranteed to find the best feature subset1
RFE drops the least important feature by the refitted model's importances; backward selection drops the one whose removal scores best2
Exhaustive search over 10 features must score $2^{10} = 1024$ subsets3
Search-and-score methods only work with models that provide feature importances4

Timed test