MOHAQ: Multi-Objective Hardware-Aware Quantization of recurrent neural networks.
Nesma M. RezkTomas NordströmDimitrios StathisZain Ul-AbdinEren Erdal AksoyAhmed HemaniPublished in: J. Syst. Archit. (2022)
Keyphrases
- recurrent neural networks
- multi objective
- optimization algorithm
- evolutionary algorithm
- low cost
- neural network
- hardware and software
- genetic algorithm
- feed forward
- artificial neural networks
- multi objective optimization
- objective function
- complex valued
- reservoir computing
- particle swarm optimization
- real time
- echo state networks
- recurrent networks
- feedforward neural networks
- long short term memory
- neural model
- hardware implementation
- conflicting objectives
- multiple objectives
- nsga ii
- quantization error
- embedded systems
- computer systems
- pareto optimal
- neural models
- data mining
- chaotic time series
- cascade correlation
- long term