Performance-Portable Autotuning of OpenCL Kernels for Convolutional Layers of Deep Neural Networks.
Yaohung M. TsaiPiotr LuszczekJakub KurzakJack J. DongarraPublished in: MLHPC@SC (2016)
Keyphrases
- deep belief networks
- neural network
- deep learning
- unsupervised learning
- multi layer
- multiple layers
- single layer
- restricted boltzmann machine
- probabilistic model
- hierarchical models
- network architecture
- fuzzy logic
- generative model
- kernel methods
- unsupervised feature learning
- feature space
- pattern recognition
- artificial neural networks
- sparse coding
- fuzzy systems
- neural network model
- support vector
- self organizing maps
- kernel function
- lightweight
- genetic algorithm
- shared memory
- machine learning
- multiple kernel learning
- feature maps
- feed forward
- back propagation
- real time
- digital signal processing
- activation function
- gaussian processes
- multilayer perceptron
- parallel algorithm