The support vector machine learning approach

$1$ point

learns a quadratic decision boundary

learns a linear decision boundary

is a classifier whose classes are identified through either positive or negative values and is trained with a cost function known as hinge loss