Efficient Sparse Neural Networks Using Regularized Multi Block Sparsity Pattern on a GPU.
Dharma Teja VooturiKishore KothapalliPublished in: HiPC (2019)
Keyphrases
- neural network
- pattern recognition
- sparse representation
- sparsity constraints
- high dimensional
- real time
- sparsity inducing
- sparse matrices
- pattern matching
- multilayer perceptron
- multi layer
- parallel architectures
- mixed norm
- sparse linear
- graphics hardware
- compressed sensing
- parallel implementation
- neural nets
- parallel processing
- general purpose
- least squares
- fuzzy logic