PARALLELIZING MATRIX FACTORIZATION ACROSS HARDWARE ACCELERATORS
A computer-implemented method includes receiving a rating matrix R. A matrix X is calculated in a matrix factorization of R, where R≈X·ΘT. Calculating X includes selecting a first value for variable p and a second value for variable q; partitioning ΘT by columns into p partitions of ΘT; partitioning...
Gespeichert in:
Hauptverfasser: | , |
---|---|
Format: | Patent |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | A computer-implemented method includes receiving a rating matrix R. A matrix X is calculated in a matrix factorization of R, where R≈X·ΘT. Calculating X includes selecting a first value for variable p and a second value for variable q; partitioning ΘT by columns into p partitions of ΘT; partitioning X by rows into q partitions of X; and partitioning R by rows and columns into p*q partitions of R. Calculating X further includes copying to each accelerator, of a plurality of accelerators, a corresponding partition of ΘT, as well as a partition of R corresponding to the accelerator and to a current partition of X. Calculating X further includes calculating, by the plurality of accelerators, a plurality of partial solutions for the current partition of X, and aggregating the plurality of partial solutions into a solution for the current partition of X. |
---|