Math 77B: Collaborative Filtering

netflix logo

     Many companies collect data at an unprecedented scale. Online stores such as Amazon.com collect clicking patterns of people navigating their webpages, credit scoring companies such as Experian and banks record people's financial histories, Netflix records people's interest in movies, and so on.
     A new field is starting to emerge known as "collaborative filtering" where this type of data is used to predict quantities of interest: What is the next book a customer would buy? Will this person pay his/her loan?, What are the next movies this customer will be interested in?
     As evidence for the prominence of this problem in industry, Netflix announced a challenge two years ago where anyone who could improve their customer recomemndation system by more that 10% would receive $1,000,000.
     This course will be based around the Netflix dataset and other collaborative filtering data sets. Students will study the theoretical aspects of clustering algorithms, matrix factorizations, and statistical estimization in order to approach the collaborative filtering application problems.