Predicting statistics of asynchronous SGD parameters for a large-scale distributed deep learning system on GPU supercomputers.
Yosuke OyamaAkihiro NomuraIkuro SatoHiroki NishimuraYukimasa TamatsuSatoshi MatsuokaPublished in: IEEE BigData (2016)
Keyphrases
- deep learning
- unsupervised learning
- unsupervised feature learning
- commodity hardware
- machine learning
- parallel computing
- mental models
- weakly supervised
- massively parallel
- expectation maximization
- learning strategies
- text classification
- learning algorithm
- data sets
- natural images
- maximum likelihood
- conditional random fields
- graphical models
- viewpoint
- bayesian networks
- clustering algorithm
- computer vision
- data mining