Note: These instructions apply to Udacity students who are enrolled in the Machine Learning 2: Unsupervised Learning course. They do not apply to Georgia Tech OMSCS students.
Recommendation Systems are an increasingly popular application of Machine Learning in many industries. If you’ve used services like Amazon, Netflix or Pandora, you might notice their personalized recommendations suggesting items for you to buy, movies to watch, or songs to listen to.
An example of how influential a well-designed recommendation system can be is what happened to Joe Simpson's Touching the Void, a book about mountain-climbing. The book was not a bestseller when it was published originally in 1988, but surged in popularity after another mountain-climbing book, Into Thin Air by Jon Krakauer, topped the charts in 1997. This happened because Amazon’s recommendation system noticed that a few people who bought and enjoyed Into Thin Air also bought and enjoyed Touching the Void. Without good online recommendation systems to suggest the book to readers, Touching the Void may have been quickly forgotten. Instead, it became very popular in its own right and in some respects exceeded the popularity of Into Thin Air.
Similarly, in this project, we will recommend movies to users. We will use the MovieLens dataset (instructions to download the dataset are in the project instructions) to create a recommendation engine like Netflix’s. We will then use the Clustering Algorithms we have learned earlier in the course to refine our engine to make better and better recommendations.
Are you ready to recommend movies to me? The following sections will get you started!