In the last few lessons,
you've been building a binary classifier. One which detects two different types
of objects, horse or human, cat or dog, that type of thing. In this lesson, we'll take a look at how
we can extend that for multiple classes. Remember when we were
classifying horses or human, we had a file structure like this. There were subdirectories for each class,
where in this case we only had two. The first thing that you'll need
to do is replicate this for multiple classes like this. It's very similar and here you can
see that both the training and validation have three subdirectories. One for Rock, one for Paper,
and one for Scissors. In these, we can put training and
validation images for Rock, Paper, and Scissors.