A New Feature

Question:

Start Quiz:

Solution:

INSTRUCTOR NOTE:

Just to clarify--the labels are outputs when testing, or making predictions. When you're training, of course you need to provide the labels to the algorithm to help it figure out the patterns, which might make the labels sound like inputs in a way.

The main point here is that there's a feature transformation going on in the SVM, which will change the way you think about "linearly separable" data.

Next Concept