DocumentCode :
1696231
Title :
Computing with Activities V. experimental proof of the stability of closed self organizing maps (gSOMs) and the potential formulation of neural nets
Author :
Reuter, Matthias
Author_Institution :
Cutec Inst. GmbH
fYear :
2008
Firstpage :
1
Lastpage :
6
Abstract :
In the last years the method of the computing of activities (CWA) has found it application in several industrial and medical applications. So we can show that especially predictive control can by handled by this special coding of information on closed self organizing maps (gSOMs). Nevertheless until today the mathematical proof, why CWA can guarantee predictive control and especially can guaranty stability was outstanding. The following paper presents the experimental proof of these premises, if a potential oriented formulation of the neural nets is used. Furthermore we show that the parameters impulse and impact of the nets leads to a better control of SOMs.
Keywords :
self-organising feature maps; closed self organizing maps stability; neural impact; neural impulse; neural nets; predictive control; Biomedical equipment; Computer industry; Frequency; Medical services; Multidimensional systems; Neural networks; Neurons; Predictive control; Self organizing feature maps; Stability; Computing with Activities; Neural Impact; Neural Impulse;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Automation Congress, 2008. WAC 2008. World
Conference_Location :
Hawaii, HI
Print_ISBN :
978-1-889335-38-4
Electronic_ISBN :
978-1-889335-37-7
Type :
conf
Filename :
4699032
Link To Document :
بازگشت