Skip to content

Introduction to Recommender Systems

September 10, 2013

This is another Coursera class I’m taking on Recommender Systems. This subject matter has always interested me and it dovetails with my interest in machine learning and data science in general. This class is provided by the University of Minnesota with two tracks: concepts and programming. Programming is a superset of Concepts. This class is significantly longer than the Coursera classes I’ve taken previously: this is a 14 week class.

The first week was a very gentle overview of recommender systems and how to recognize them in the wild. The homework assignments for the first week were to take a movie rating survey to provide a data source for later projects, and finding and analyzing a recommender used in the wild.

The second week was an explanation of non-personalized recommenders. These recommenders use aggregate data points to make the same recommendations to all viewers. Various formulas for providing recommendations were provided and used in the written homework #2. In the assignment a movie data set was provided and we submitted 4 ways to recommend the top 5 movies:

  1. the mean (average) of all the ratings of each user for a movie.
  2. the percentage of ratings that were 4+ (on a 1-5 scale). So counting the number of ratings with 4 or 5 and dividing that number by the total number of rating received for that movie.
  3. a simple count of the number of ratings (whatever the score) provided for a movie.
  4. Movies most like Star Wars: Episode 4 — this use a count of ratings from users only if they also rated Star Wars Episode 4 and divided that by the total number of ratings for that film. A way of providing a similarity between movies: the higher this number the more people had rated (and presumably watched) both movies.

 

Advertisements

From → General

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: