A Functional Manipulation for Improving Tolerance against Multiple-Valued Weight Faults of Feedforward Neural Networks.
Naotake KamiuraYasuyuki TaniguchiNobuyuki MatsuiPublished in: ISMVL (2001)
Keyphrases
- feedforward neural networks
- multiple valued
- multi valued
- neural network
- file organization
- back propagation
- recurrent neural networks
- hidden layer
- boolean functions
- multilayer perceptron
- continuous attributes
- training algorithm
- complex valued
- fault diagnosis
- activation function
- variable selection
- radial basis function
- genetic algorithm