A Learning Method for Neural Networks Based on a Pseudoinverse Technique
A theoretical formulation of a fast learning method based on a pseudoinverse technique is presented. The efficiency and robustness of the method are verified with the help of an Exclusive OR problem and a dynamic system identification of a linear single degree of freedom mass–spring problem. It is o...
Main Authors: | Chinmoy Pal, Ichiro Hagiwara, Naoki Kayaba, Shin Morishita |
---|---|
Format: | Article |
Language: | English |
Published: |
Hindawi Limited
1996-01-01
|
Series: | Shock and Vibration |
Online Access: | http://dx.doi.org/10.3233/SAV-1996-3306 |
Similar Items
-
Pseudoinversion of degenerate metrics
by: C. Atindogbe, et al.
Published: (2003-01-01) -
Weight and Structure Determination Neural Network Aided With Double Pseudoinversion for Diagnosis of Flat Foot
by: Liangming Chen, et al.
Published: (2019-01-01) -
Clustering based on Principal Component Analysis and Pseudoinverse Transformation
by: Sih-Yin Shen, et al.
Published: (2008) -
CONTROL END-EFFECTOR MOVEMENT OF ROBOTIC ARM BY PSEUDOINVERSE METHOD
by: Ondrej HOCK, et al.
Published: (2019-12-01) -
Experimental Investigation of Iterative Pseudoinverse Ghost Imaging
by: Xiaofeng Lv, et al.
Published: (2018-01-01)