Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks.
Urs KösterTristan WebbXin WangMarcel NassarArjun K. BansalWilliam ConstableOguz ElibolStewart HallLuke HornofAmir KhosrowshahiCarey KlossRuby J. PaiNaveen RaoPublished in: NIPS (2017)
Keyphrases
- neural network
- pattern recognition
- training algorithm
- training process
- genetic algorithm
- e learning
- feed forward neural networks
- sensitivity analysis
- feedforward neural networks
- real time
- multi layer perceptron
- multi layer
- neural network model
- computationally expensive
- cost effective
- online learning
- fuzzy logic
- artificial neural networks
- metadata
- machine learning