Parallel Algorithm Based on Singular Value Decomposition for High Performance Training of Neural Networks.
Gabriele Maria LozitoValentina LucaferriMauro ParodiMartina RadicioniFrancesco Riganti FulgineiAlessandro SalviniPublished in: ICCS (5) (2019)
Keyphrases
- singular value decomposition
- parallel algorithm
- neural network
- training process
- pc cluster
- singular values
- least squares
- parallel computation
- parallel computers
- low rank
- dimension reduction
- principal component analysis
- latent semantic indexing
- cluster of workstations
- shared memory
- dimensionality reduction
- linear algebra
- feature selection
- training samples
- binary search trees
- training set
- distributed memory
- pattern recognition
- data sets
- computer vision
- parallel version
- medial axis transform
- singular vectors