Title :
Modification of neural network by the self-organization method
Author :
Neusipin, K.A. ; Sholohov, D.O.
Author_Institution :
Bauman Moscow State Tech. Univ., Moscow
Abstract :
Algorithm of modified neural network based on Volterra kind network with the weight coefficients formed by the self organization method was developed. Modified algorithm of neural network in comparison with usual algorithm needs less computer memory and also a shorter time period to complete, which is very important for onboard digital computers.
Keywords :
self-organising feature maps; Volterra network; neural network; self-organization method; weight coefficient; Acceleration; Automatic control; Buildings; Communication system control; Computer networks; Control systems; Finishing; Navigation; Neural networks; Testing; Algorithm; method; neural network; self-organization;
Conference_Titel :
Control and Communications, 2009. SIBCON 2009. International Siberian Conference on
Conference_Location :
Tomsk
Print_ISBN :
978-1-4244-2007-0
DOI :
10.1109/SIBCON.2009.5044840
