Unified stabilization approach to principal and minor components extraction algorithms

Neural Networks

Volumn 14, Issue 10, 2001, Pages 1377-1387

  Chen, Tianping ^a   Amari, Shun Ichi ^b  

^a FUDAN UNIVERSITY   (China)

^b RIKEN BRAIN SCIENCE INSTITUTE   (Japan)

Author keywords

Minor component extraction; Natural gradient; Principal component extraction; Singular value decomposition; Stability of dynamical system

Indexed keywords

ALGORITHMS; EXTRACTION; STABILIZATION;

PROCESSING FIELDS;

NEURAL NETWORKS;

ALGORITHM; ARTICLE; DYNAMICS; EXTRACTION; INFORMATION PROCESSING; NERVE CELL NETWORK; PRINCIPAL COMPONENT ANALYSIS; PRIORITY JOURNAL; TECHNIQUE;

ALGORITHMS; COMPUTER SIMULATION; NEURAL NETWORKS (COMPUTER);

EID: 0035187852     PISSN: 08936080     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0893-6080(01)00116-2     Document Type: Article

References (20)

1. Neural theory of association and concept-formation
2. Information geometry
3. Natural gradient works efficiently
4. Stability analysis of learning algorithms for blind source separation
5. Noholonomic orthogonal constraints in adaptive blind source separation
6. Least square matching problems
7. Dynamical systems that sort lists, diagonalize matrices, and solve linear programming problems
8. Modified Oja's algorithms for principal subspace and minor subspace extraction
9. A unified approach for principal and minor components extraction
10. Global convergence of Oja's subspace algorithm for principal component extraction
11. A self-stabilized minor subspace rule
12. The geometry of algorithms with orthogonality constraints
13. A simplified neuron model as a principal component analyzer
14. Principal components, minor components, and linear neural networks
15. On stochastic approximation of the eigenvectors and eigenvalues of the expectation of a random matrix
16. Optimal unsupervised learning in a single-layer linear feedforward network
17. A unified neural bigradient algorithm for robust PCA and MCA
18. Neural nets for dual subspace pattern recognition method
19. Least mean square error recognition principle for self-organizing neural nets