As discussed in the introduction to this section, RBMs can be used in a number of situations, in either a supervised or unsupervised manner. Collaborative filtering refers to a strategy for predicting user preferences on the underlying assumption that if user A likes item Z, and user B also likes item Z, then user B might also like something else that user A likes (say, item Y).
We see this use case in action every time Netflix recommends something to us or every time Amazon recommends us a new vacuum cleaner (because, of course, we bought a vacuum cleaner and are now very clearly into domestic appliances).
So now that we've covered a bit of theory on what RBMs are, how they work, and how they are used, let's jump into building one!