#model-based-selection

Question 1: Select all that apply

You want to keep only the most useful columns of a dataset with many features. Which statements about feature selection methods are true? Select all that apply.

Simple association-based selection, such as ranking features by their correlation with the target, ignores interactions between features1
With a linear model on scaled features, you can select features by dropping those whose coefficients are close to zero2
Recursive feature elimination removes features in the same order as one ranking of the full model's feature importances3
Model-based feature selection can use a $k$ -NN classifier to judge which features are important4

#model-based-selection

Timed test