-
Notifications
You must be signed in to change notification settings - Fork 8
Open
Description
Recently Spark added a built-in implementation of the SVD as a method on the RowMatrix class. I'm interested in running comparing the performance of this algorithm against the SVD implemented in thunder-factorization. If the Spark version is favorable, we should consider adding support for it in the factorization algorithms that use the SVD
Metadata
Metadata
Assignees
Labels
No labels