Welcome back. In this video, we're going to introduce the Capstone Project that you're going to be carrying out as the culmination of everything you've done in the recommender system specialization. In this Capstone, you're going to integrate and apply what you've learned across all four of the courses that we've seen so far into one substantial project that's going to be a lot more realistic scenario of how you would actually go about developing a plan for integrating recommendation into a real application. The goal of this is to evaluate and allow you to demonstrate the knowledge that you've accumulated across the whole set of material in an integrated fashion. For the main Capstone project, you're going to select a recommender system use case. We're going to give you several different ones to choose from. And your job is to analyze this case and develop a plan for how users might interact with the recommender, how you would evaluate what algorithms are going to perform well on this recommendation case, what some potential candidate algorithms you might want to try are, etc. And for those of you who are completing the Honors track, in addition to that main Capstone, you'll be carrying out an empirical algorithm study given a data set and a use case, though not the same one that you did in your analysis. We will give you one where we actually have a live data set and in-lens kit, you will go through the process of analyzing, developing the reports and making the recommendation on the strengths and weaknesses of particular algorithms for that particular use case. Structurally, this course is unlike the four courses that preceded. It's not filled with videos and quizzes. You're going to have instructions, criteria and then go off and work for several weeks and get the projects done. There is no incremental grading, you'll get peer grading of the entire project submission. Expect that you're also going to have to grade three or more of other people's projects, and you will get a rubric of grading criteria that make it clear particularly in the main course analysis. This is not about a correct answer, it's about a well reasoned plan that references things that you've learned, make sense against what you've learned. In the honors component which is independent and can be done before, after or in parallel, again the criteria are going to be about having made well reasoned choices about how to do the evaluation. This is not a mechanical evaluation of, well, if you have found 0.03 was your statistic, then you did it right. We're looking to see that at this point, you can take a higher level criteria and make your own decisions about exactly how to go through it and come back with something that we and your peers can look at and say, "Yeah, that means you really get it." If you get stuck, the discussion forums are there. You can use them for general discussion, general support in order to be able to get past issues with lens kit if you're in the honors track, or to clarify what's going on in the requirements for the main component of the class. One last word before we set you off on this, Capstone projects educationally have a very specific purpose, and that purpose is for you to pull together a bunch of things that you've learned, not only to demonstrate them but to solidify your knowledge. We want to encourage you, don't throw away what you have already. These projects are going to be an ideal time to go back, take a look at your notes, the lecture slides, whatever it is in the previous four courses because that's where you're going to find all the information that you need to successfully complete this project. We look forward to seeing you at the end.
