[ml] video on sparse SVD in R

[Mike Schachter]

Saw a nice post on reddit ML to a video which
talks about an SVD package used by the winning
netflix prize team:

http://www.r-bloggers.com/the-netflix-prize-big-data-svd-and-r/

SVD is singular-value decomposition, it decomposes a
non-square matrix into eigenvalues and eigenvectors:

http://en.wikipedia.org/wiki/Singular_value_decomposition

The video is about the irlba package, which provides an
efficient way to do SVD on large sparse matricies:

http://www.rforge.net/irlba/