Networks with Negative Edges


When I started my PhD program at the Technische Universität in Berlin, my first task was to research rating prediction methods.  Rating prediction is a huge topic:  In 2006, Netflix started the Netflix Prize, in which participants could win a million dollars by predicting whether a given person likes or dislikes a given movie.  To make these predictions possible, Netflix published the ratings of 480,000 persons for 17,000 movies.  Ratings on Netflix are on a scale from 1 to 5.  1 would mean a user dislikes a movie and 5 means the user likes it.  Then, given a list of (user, movie) pairs, ratings had to be predicted and submitted to Netflix.  The submitted predictions were then compared to actual withheld ratings, and the first team to improve by 10% the prediction accuracy (measured in a least-squares sense) of Netflix’s own recommender then won the million dollars.

To predict…

View original post 2,131 more words


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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s