First-order approximation of Gram-Schmidt orthonormalization beats deflation in coupled PCA learning rules.
Ralf MöllerPublished in: Neurocomputing (2006)
Keyphrases
- learning rules
- principal component analysis
- neural network
- biologically inspired
- hybrid neural
- artificial neural networks
- taylor series
- feature extraction
- dimensionality reduction
- principal components analysis
- learning scheme
- principal components
- higher order
- back propagation
- error bounds
- boltzmann machine
- approximation algorithms
- face recognition
- first order logic
- covariance matrix
- principle component analysis
- learning algorithm
- independent component analysis
- feature space
- real time
- multi modal
- linear discriminant analysis
- image quality
- co occurrence
- feature selection
- fuzzy logic
- image analysis
- machine learning
- three dimensional
- learned rules
- data mining
- pattern recognition