Title :
Sigmoid-Linear BP Neural Network Based Variable Selecting Method
Author_Institution :
Dept. of Civil Eng., Sichuan Coll. of Archit. Technol., Deyang, China
Abstract :
Variable selection is a very vital step of models making and prediction researching. In order to solve the problem of variable selection in complicated system, statistical and neural network theory are combined to propose a method of scatter determine quotiety based on Sigmoid-Linear BP neural network in this paper. The efficiency of this method is also proved by simulation studies.
Keywords :
backpropagation; neural nets; statistical analysis; backpropagation; sigmoid-linear BP neural network; statistical theory; variable selecting method; Cities and towns; Civil engineering; Computer architecture; Computer networks; Educational institutions; Input variables; Neural networks; Predictive models; Reactive power; Scattering; BP neural network; determine quotiety; scatter; variable selection;
Conference_Titel :
Natural Computation, 2009. ICNC '09. Fifth International Conference on
Print_ISBN :
978-0-7695-3736-8
DOI :
10.1109/ICNC.2009.410
