A Power-Aware Digital Multilayer Perceptron Accelerator with On-Chip Training Based on Approximate Computing.
Duckhwan KimJaeha KungSaibal MukhopadhyayPublished in: IEEE Trans. Emerg. Top. Comput. (2017)
Keyphrases
- multilayer perceptron
- backpropagation algorithm
- feedforward neural networks
- back propagation
- neural network
- artificial neural networks
- radial basis function
- hidden layer
- radial basis function network
- training algorithm
- feed forward neural networks
- neural network model
- probabilistic neural networks
- multilayer neural network
- commonly applied
- circuit design
- multi layer perceptron
- low cost
- input vector
- high speed
- chip design
- input vectors
- power consumption
- modular neural network
- feed forward
- power management
- neuro fuzzy models
- training set
- ibm power processor
- activation function
- rbf neural network
- mixed signal
- data sets
- phase locked loop
- power dissipation
- parallel implementation
- embedded systems
- support vector machine svm
- control system
- reinforcement learning
- feature extraction