Question $4$

The cost function in the Support Vector Machine

$1$ point

ensures that each training example is on the decision boundary

penalizes errors on query examples

ensures that each example in a linearly separable training set is on the right side of the decision boundary

$($select all that apply$)$