Mixed-precision training of deep neural networks using computational memory.
S. R. NandakumarManuel Le GalloIrem BoybatBipin RajendranAbu SebastianEvangelos EleftheriouPublished in: CoRR (2017)
Keyphrases
- online learning
- neural network
- computational power
- training process
- pattern recognition
- computing power
- neural network model
- auto associative
- associative memory
- training algorithm
- feedforward neural networks
- back propagation
- multi layer perceptron
- training patterns
- neural network training
- neural network structure
- fuzzy systems
- neural nets
- multilayer perceptron
- learning algorithm
- artificial neural networks
- fault diagnosis
- self organizing maps
- high precision
- memory requirements
- fuzzy neural network
- main memory
- average precision
- network architecture
- memory size
- feed forward
- test set
- backpropagation algorithm
- precision and recall
- data sets
- error back propagation