Accelerating divergent applications on SIMD architectures using neural networks.
Beayna GrigorianGlenn ReinmanPublished in: ICCD (2014)
Keyphrases
- neural network
- parallel architectures
- single instruction multiple data
- pattern recognition
- parallel algorithm
- artificial neural networks
- neural architectures
- back propagation
- multilayer perceptron
- genetic algorithm
- fuzzy logic
- training process
- training algorithm
- multi layer
- self organizing maps
- fuzzy neural network
- multi layer perceptron
- information systems
- parallel processing
- processor array
- highly parallel
- competitive learning
- activation function
- parallel implementation
- fuzzy systems
- neural nets
- associative memory
- recurrent neural networks
- feed forward