Title :
A modular realization of adaptive PCA
Author :
Shirazi, Malltlad S. ; Noda, Hideki ; Sawai, Hidefumi
Author_Institution :
Commun. Res. Lab., Minist. of Posts & Telecommun., Japan
Abstract :
We propose an adaptive PCA algorithm which alleviates suboptimality of the PCA method for nonstationary signals. A modular neural realization of adaptive PCA is considered and its design is formulated as an optimization problem, following the design of the vector quantizer. This formulation results in a competitive algorithm that learns data´s local eigenstructures in an unsupervised way. The algorithm includes the recently proposed adaptive transform coding algorithm of R.D. Dony and S. Haykin (1995) as a special case and, as confirmed by simulation studies, the algorithm is better than their algorithm in mean square error (MSE)
Keywords :
adaptive systems; competitive algorithms; eigenvalues and eigenfunctions; neural nets; statistical analysis; unsupervised learning; vector quantisation; MSE; adaptive PCA algorithm; adaptive transform coding algorithm; competitive algorithm; local eigenstructures; mean square error; modular neural realization; modular realization; nonstationary signals; optimization problem; principal component analysis; simulation studies; statistical technique; suboptimality; unsupervised learning; vector quantizer; Data compression; Decorrelation; Feature extraction; Laboratories; Mean square error methods; Principal component analysis; Random processes; Stability; Statistics; Transform coding;
Conference_Titel :
Systems, Man, and Cybernetics, 1997. Computational Cybernetics and Simulation., 1997 IEEE International Conference on
Print_ISBN :
0-7803-4053-1
DOI :
10.1109/ICSMC.1997.633055
