Hardware-Based Principal Component Analysis for Hybrid Neural Network Trained by Particle Swarm Optimization on a Chip.
Tuan Linh DangYukinobu HoshinoPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2019)
Keyphrases
- hybrid neural network
- particle swarm optimization
- principal component analysis
- low cost
- vlsi implementation
- host computer
- single chip
- pso algorithm
- programmable logic
- circuit design
- hardware and software
- principal components
- neural network
- particle swarm optimization algorithm
- training procedure
- multi objective
- ibm power processor
- independent component analysis
- evolvable hardware
- ibm zenterprise
- swarm intelligence
- chip design
- high speed
- differential evolution
- global optimization
- covariance matrix
- training set
- image sensor
- face recognition
- feature space
- particle swarm optimization pso
- real time
- dimensionality reduction
- processor core
- low power consumption
- low power
- linear discriminant analysis
- genetic algorithm
- face images
- massively parallel
- feature extraction
- hardware implementation
- field programmable gate array
- embedded systems
- analog vlsi
- multithreading
- digital circuits
- support vector machine
- computer systems