A new variable shape parameter strategy for RBF approximation using neural networks.
Fatemeh Nassajian MojarradMaria Han VeigaJan S. HesthavenPhilipp ÖffnerPublished in: Comput. Math. Appl. (2023)
Keyphrases
- neural network
- radial basis function
- artificial neural networks
- shape descriptors
- pattern recognition
- shape model
- radial basis function network
- back propagation
- basis functions
- rbf network
- parameter space
- approximation algorithms
- approximation error
- rule extraction
- multi layer perceptron
- single parameter
- function approximation
- mlp neural networks
- shape analysis
- error bounds
- fuzzy logic
- shape features
- shape representation
- rbf neural network
- support vector machine svm
- maximum likelihood estimates
- radial basis function neural network
- genetic algorithm
- recurrent neural networks
- fuzzy artmap
- feature extraction
- polygonal approximation
- training set
- support vector machine
- activation function
- closed form
- neural nets
- parameter settings