Since October 2006, Netflix has held a competition to improve their movie recommendation system.  The Netflix Prize is $1 million for anyone who can beat their existing system with a 10% accuracy improvement.  So far, around 30,000 teams have registered for the challenge and the top ten teams boast improvements around 9% over the incumbent recommendation engine, Cinematch.  Each team of competitors is hard at work generating an algorithm that can predict human behavior through movie predictions.  However, many believe they’ve reached the limits of computer prediction. Applying mathematical formulas to the data set provided by Netflix has yielded impressive improvements and the New York Times Magazine article does a great job in breaking down the equations:

This content is available for Premium Subscribers only.
Already a subscriber? Log in