How do you decide netflix prize?

So, I just pound into the wild. I actually don't really know much about data. I ask out of sheer interest, because I really will not have time to try to participate in this competition.

But just for fun, how would you handle this?

It works something like this: you get a really large set of movie IDs and user votes. Now I got a few votes from a user and a film, what rating will he give in this film?

EDIT URL for the specified prize http://www.netflixprize.com/

+3
source share
4 answers

, , , :)

, , . , , ( ), 20% 80% .

, Napoleon Dynamite, . , , , .. , "" , , -, , , , , , .

+1

, :

. , .. . , .

, , ​​, :

= + .

.

: - node, , .

. "" , .

:

= + + .

, .

: 5 . 5- . , 4 5 . .

:

= + + 5 + 4-5 +... + 5-4-3-2-1

!

+1

So, perhaps for those 3 readers who are NOT completely familiar with linear regression, like me: they demand to improve their forecasts by 10%. It's hard. This is difficult because I believe that rating a user’s choice is simply the average choice that other users gave is probably a good rating. I want to say that there is not much space left for improvement.

-1
source

All Articles