Title :
A self-organizing neural net with three functions related to principal component analysis
Author :
Matsuoka, Kiyotoshi ; Kawamoto, Mitsuru
Author_Institution :
Dept. of Control Eng., Kyushu Inst. of Technol., Kitakyushu, Japan
Abstract :
The most remarkable property of the authors´ model is that the three mathematical functions related to principal component analysis (principal component analyzer, orthogonal projection operation, and novelty filter) are acquired by the same network but with different learning parameters
Keywords :
learning (artificial intelligence); self-organising feature maps; stochastic processes; novelty filter; orthogonal projection operation; principal component analysis; principal component analyzer; self-organizing neural net; Control engineering; Eigenvalues and eigenfunctions; Filters; Mathematical model; Matrix decomposition; Neural networks; Principal component analysis; Stochastic processes;
Conference_Titel :
Systems, Man and Cybernetics, 1993. 'Systems Engineering in the Service of Humans', Conference Proceedings., International Conference on
Conference_Location :
Le Touquet
Print_ISBN :
0-7803-0911-1
DOI :
10.1109/ICSMC.1993.384941
