Netflix Progress Prize Awarded

As per the prize rule, they are under the obligation to describe and publish their mode (as they did last time). I really wish someone would release an open source recommender package, coded in C, with PHP bindings, based on the latest / greatest algorithms, or any algorithm (just plain SVD if anything).<p>Interesting links (from the winning team(s) websites about their latest research) : <a href="http://public.research.att.com/~volinsky/netflix/kdd08koren.pdf" rel="nofollow">http://public.research.att.com/~volinsky/netflix/kdd08koren....</a> <a href="http://www.commendo.at/index.php?lang=0&#38;_0=2&#38;_1=3" rel="nofollow">http://www.commendo.at/index.php?lang=0&#38;_0=2&#38;_1=3</a>

It may be simply mathematically impossible to get to 10 percent given the observable variables. Maybe netflix knows this and just gets synthesized research on the cheap.

I started working on this about a month and a half ago. My RMSE is pathetic thus far(just started tweaking a kNN with Pearson correlation as a distance metric), but I don't really care. It's a great way to brush up on a lot of CompSci concepts at once you may be rusty on. Besides obviously machine learning, when you're trying to process 2 gb of training files on your personal computer, O(n) time/space complexity REALLY starts to matter, as does choice of implementation language...using mostly Python with C extensions for the heavy math right now...

I'd love to win just so I could settle arguments by saying "anyone who has won a million dollar data mining competition raise your hand"

I keep hoping someone will swoop down with a crazy solution and snatch the prize from the people trying to solve it conventionally. So much more romantic than this slog!

2 years in and they still haven't given out the prize money. How's that crowd sourcing working out for you NetFlix?