Stability of Oja's PCA subspace rule
Neural Computation
Principal component neural networks: theory and applications
Principal component neural networks: theory and applications
Random Iterative Models
Handwritten Digit Recognition by Local Principal Components Analysis
ISMIS '97 Proceedings of the 10th International Symposium on Foundations of Intelligent Systems
Oja's algorithm for graph clustering, Markov spectral decomposition, and risk sensitive control
Automatica (Journal of IFAC)
Stability Analysis of Oja-RLS Learning Rule
Fundamenta Informaticae
Hi-index | 0.00 |
By elementary tools of matrix analysis, we show that the discrete dynamical system defined by Oja algorithm is stable in the ball K(0, 81=64) if only gains βn are bounded by (2B)-1; where B = b2 and b is the bound for the learning sequence. We also define a general class of Oja's systems (with gains satisfying stochastic convergence conditions) which tend to the infinity with exponential rate if only their initial states are chosen too far from the zero point.