Power-Performance Evaluation of Parallel Multi-objective EEG Feature Selection on CPU-GPU Platforms.
Juan José EscobarJulio OrtegaAntonio Francisco DíazJesús GonzálezMiguel DamasPublished in: ICA3PP (2017)
Keyphrases
- multi objective
- graphics processing units
- feature selection
- multithreading
- general purpose
- parallel computing
- parallel processing
- parallel implementation
- gpu implementation
- optimization algorithm
- multi objective optimization
- parallel computation
- pc cluster
- parallel programming
- graphics processors
- evolutionary algorithm
- cluster of workstations
- computational power
- multi threaded
- distributed memory
- multiple objectives
- power consumption
- massively parallel
- memory bandwidth
- genetic algorithm
- graphics hardware
- floating point
- pareto optimal
- real time
- shared memory
- heterogeneous computing
- text categorization
- mutual information
- conflicting objectives
- multi objective optimization problems
- compute unified device architecture
- high performance computing
- level parallelism
- data transfer
- signal processing
- particle swarm optimization
- multi objective evolutionary
- feature space
- nsga ii
- objective function
- feature selection algorithms
- graphic processing unit
- processing units
- efficient implementation
- text classification
- support vector machine
- classification accuracy
- support vector
- multi core processors
- eeg signals
- motor imagery
- brain computer interface
- computing systems
- feature subset
- parallel hardware
- feature set
- neural network