Scalable energy-efficient, low-latency implementations of trained spiking Deep Belief Networks on SpiNNaker.
Evangelos StromatiasDaniel NeilFrancesco GalluppiMichael PfeifferShih-Chii LiuSteve B. FurberPublished in: IJCNN (2015)
Keyphrases
- low latency
- energy efficient
- deep belief networks
- wireless sensor networks
- energy consumption
- highly efficient
- sensor networks
- high speed
- unsupervised learning
- virtual machine
- real time
- high throughput
- energy efficiency
- training set
- probabilistic model
- base station
- restricted boltzmann machine
- deep learning
- generative model
- multilayer perceptron
- mobile nodes
- routing protocol
- feed forward
- stream processing
- data center
- routing algorithm
- machine learning
- neural network
- mobile agents
- operating system