The Fisher Information as a Neural Guiding Principle for Independent Component Analysis.
Rodrigo EchevesteSamuel EckmannClaudius GrosPublished in: Entropy (2015)
Keyphrases
- independent component analysis
- fisher information
- free energy
- principal component analysis
- independent components
- signal processing
- factor analysis
- noise level
- blind source separation
- neural network
- euclidean space
- probability density function
- positive definite
- multidimensional scaling
- estimation error
- linear discriminant analysis
- probability distribution
- feature space
- machine learning