Netflix, Dark Knowledge, and Why Simpler Can Be Better

Weary from an all-night coding effort, and rushed by the looming 6:42PM deadline, Lester Mackey searched franticly for the proper prediction file to submit. Lester was a member of “The Ensemble”—a large coalition of data scientists who had joined forces late in the great Netflix challenge, after a grand prize of $1 Million — not to mention substantial prestige in the data science community. After 2 years of effort and frequent lead changes (lately, mostly with rival mash-up team “Belkor’s Pragmatic Chaos”), Lester’s team looked to have finally achieved its goal…. Rewind about 2 years to October 2006: Netflix unleashed its challenge to crowd source one of its most difficult challenges: making quality movie recommendations to its customer base. The competition goal was to use a customers’ previous movie reviews…


Link to Full Article: Netflix, Dark Knowledge, and Why Simpler Can Be Better