Accelerating Matrix Multiplication in Deep Learning by Using Low-Rank Approximation.
Kazuki OsawaAkira SekiyaHiroki NaganumaRio YokotaPublished in: HPCS (2017)
Keyphrases
- deep learning
- matrix multiplication
- low rank approximation
- low rank
- singular value decomposition
- matrix factorization
- message passing
- unsupervised learning
- distributed memory
- machine learning
- collaborative filtering
- semi supervised
- multiscale
- image processing
- belief propagation
- mean shift
- missing data
- clustering algorithm
- data sets