A Methodology for Neural Network Architectural Tuning Using Activation Occurrence Maps.
Rafael GarciaAlexandre Xavier FalcãoAlexandru C. TeleaBruno Castro da SilvaJim TørresenJoão Luiz Dihl CombaPublished in: IJCNN (2019)
Keyphrases
- neural network
- pattern recognition
- feature maps
- feed forward neural networks
- back propagation
- artificial neural networks
- fuzzy logic
- self organizing maps
- network architecture
- multi layer
- neural network model
- fault diagnosis
- information processing
- artificial intelligence
- genetic algorithm
- parameter tuning
- neural network is trained
- training algorithm
- hidden layer
- neural nets
- parameter settings
- decision trees
- data mining