FPGA architecture for feed-forward sequential memory network targeting long-term time-series forecasting.
Kentaro OrimoKota AndoKodai UeyoshiMasayuki IkebeTetsuya AsaiMasato MotomuraPublished in: ReConFig (2016)
Keyphrases
- feed forward
- long term
- artificial neural networks
- recurrent networks
- neural architecture
- back propagation
- output layer
- neural nets
- spiking neural networks
- short term
- network architecture
- neural network
- recurrent neural networks
- hidden layer
- spiking neurons
- dedicated hardware
- peer to peer
- hardware implementation
- hardware design
- biologically plausible
- high speed
- visual cortex
- feed forward neural networks
- application level
- hardware architecture
- hardware architectures
- real time
- software implementation
- data transfer
- data flow
- parallel architecture
- activation function
- multi layer
- random access
- hardware software
- hebbian learning
- biologically inspired
- genetic algorithm