Randomized algorithms to update partial singular value decomposition on a hybrid CPU/GPU cluster.
Ichitaro YamazakiJakub KurzakPiotr LuszczekJack J. DongarraPublished in: SC (2015)
Keyphrases
- singular value decomposition
- randomized algorithms
- graphics processing units
- lower bound
- singular values
- approximation algorithms
- least squares
- dimensionality reduction
- low rank
- clustering algorithm
- latent semantic indexing
- principal component analysis
- practical problems
- worst case
- data clustering
- upper bound
- randomized algorithm
- data points
- pattern recognition
- machine learning
- multi class
- np hard
- support vector
- image sequences
- learning algorithm