Adaptive runtime exploiting sparsity in tensor of deep learning neural network on heterogeneous systems.
Kuo-You PengSheng-Yu FuYu-Ping LiuWei-Chung HsuPublished in: SAMOS (2017)
Keyphrases
- deep learning
- neural network
- heterogeneous systems
- unsupervised learning
- machine learning
- mental models
- web services
- distributed architecture
- pattern recognition
- computing systems
- weakly supervised
- higher order
- application developers
- parallel architectures
- image segmentation
- feature space
- high dimensional
- dimensionality reduction
- parallel execution
- real world