Comments on: The Netflix Prize and Production Machine Learning Systems: An Insider Look https://blogs.mathworks.com/loren/2015/04/22/the-netflix-prize-and-production-machine-learning-systems-an-insider-look/?s_tid=feedtopost Loren Shure is interested in the design of the MATLAB language. She is an application engineer and writes here about MATLAB programming and related topics. Mon, 04 Oct 2021 12:31:23 +0000 hourly 1 https://wordpress.org/?v=6.2.2 By: Toshi Takeuchi https://blogs.mathworks.com/loren/2015/04/22/the-netflix-prize-and-production-machine-learning-systems-an-insider-look/#comment-44965 Fri, 24 Apr 2015 03:33:56 +0000 https://blogs.mathworks.com/loren/?p=1159#comment-44965 Thanks Brad, I am glad you like it. I didn’t cover it here, but another interesting technique that emerged from the Netflix Prize was Matrix Factorization/SVD, very similar to LSA that you saw in my previous guest post “Can You Find Love through Text Analytics?”, where I used SVD to derive a low rank approximation of the document term matrix to cluster similar profiles. Instead of words counts or TFIDF, you can use the movie ratings to represent the matrix. Only issue is that you can’t use SVD when the matrix has any missing values, but the whole point of recommer systems is to predict the missing values. To overcome this issue and reduce heavy computational demand of SVD, number of clever techniques were used in the Netflix Competition, such as FunkSVD. Interesting stuff.

]]>
By: Brad Stiritz https://blogs.mathworks.com/loren/2015/04/22/the-netflix-prize-and-production-machine-learning-systems-an-insider-look/#comment-44964 Thu, 23 Apr 2015 02:50:43 +0000 https://blogs.mathworks.com/loren/?p=1159#comment-44964 Toshi, thank you for another outstanding blog post. Very well-researched and well-written, and extremely interesting!

]]>