Title :
A study of neural network architecture for weak non-linear modeling
Author :
Mizukami, Y. ; Wakasa, Yuji ; Tanaka, Kanya
Author_Institution :
Yamaguchi Univ., Ube, Japan
Abstract :
This paper studies a property of neural network architecture for non-linear modeling. This method was proposed in our previous work and has three improvements; 1) the design of a sigmoidal function with localized derivative, 2) a deterministic scheme for weight initialization, and 3) an updating rule for weight parameters. We discuss its robustness against noise based on simulation results.
Keywords :
control nonlinearities; learning (artificial intelligence); neural net architecture; noise; stability; deterministic scheme; learning algorithm; neural network architecture; robustness against noise; sigmoidal function; updating rule; weak non-linear modeling; weight initialization; weight parameter;
Conference_Titel :
SICE 2004 Annual Conference
Conference_Location :
Sapporo
Print_ISBN :
4-907764-22-7
