Boosting irregular array Reductions through In-lined Block-ordering on fast processors.
Jan CieskoSergi MateoXavier TeruelVicenç BeltranXavier MartorellJesús LabartaPublished in: HPEC (2015)
Keyphrases
- rows and columns
- parallel algorithm
- irregularly shaped
- parallel processing
- processor array
- feature selection
- learning algorithm
- multiprocessor systems
- high end
- boosting algorithms
- efficient computation
- image blocks
- block size
- ensemble learning
- single processor
- block wise
- partial order
- boosting framework
- machine learning
- weak classifiers
- ensemble methods
- parallel architecture
- cost sensitive
- linear combination