Origin and elimination of two global spurious attractors in Hopfield-like neural network performing Boolean factor analysis.
Alexander A. FrolovDusan HúsekIgor P. MuravievPavel PolyakovPublished in: Neurocomputing (2010)
Keyphrases
- factor analysis
- neural network
- component analysis
- independent component analysis
- associative memory
- discriminant analysis
- classification and regression trees
- hopfield network
- hopfield neural network
- auto associative
- cluster analysis
- correspondence analysis
- neural network model
- matrix factorization
- statistical tests
- information criterion
- automatic model selection
- latent factors
- maximum likelihood
- pattern recognition
- similarity measure
- decision trees
- computer vision