Objective: The original goal of my project was to analyze Netflix movies by their ratings, generate statistics (about the movies), and apply machine learning techniques to build and evaluate a predictive model. When building an analysis on the data, I decided for reasons described below, to tackle 2 types of recommender systems, one given no data on a user and the other given some data on a user.
Reason for Project: I am an avid Netflix user and monthly customer. Knowing the science behind such a large company seems beneficial for one scientist to know and be able to explain.
Data Background: The data from the project is from 2009 and was originally used for a contest held by Netflix. Netflix crowd sourced development a couple of times for the best recommender system with a $100K-$1M prize. Unfortunately on the second attempt of the contest in 2010, Netflix was ligated for indirectly releasing user information to the public.