Hacker News Reader

All algorithms are interesting depending on the problem you are trying to solve.

I'd really love to have a C implementation of distributed SVD computation for large matrices.

Is the key word here "distributed?" I'm neither a C nor an ML expert, but I'd assume there must be C implementations for SVD computation.

Searching for "C support vector machine mpi" finds πSvM: "πSvM is a parallel implementation of the Support Vector Machine (SVM) algorithm that allows efficient training and testing on a multiprocessor system" [1] . From a casual inspection of the code it seems there is some C++ and C.

No idea if this is any good, but hey, it exists.

[1] - pisvm.sourceforge.net

Edit: Whoops! of course SVD != SVM. Acronyms, acronyms.

https://code.google.com/p/redsvd/

this isn't distributed but claims to be good for large matrices

Reply on news.ycombinator.com