Title : 
Neural Network Approach for Solving Boundary Value Problems of Electrodynamics by the R-Functions Method
         
        
        
            Author_Institution : 
Bauman Moscow State Tech. Univ., Moscow
         
        
        
        
        
        
        
            Abstract : 
Application of the RFM GSSs allows one to increase efficiency of training procedures for RBF neural networks. The technique proposed can be applied to solving a wide class of electrodynamics problems, including eigenvalue and nonlinear problems.
         
        
            Keywords : 
computational electromagnetics; eigenvalues and eigenfunctions; electrical engineering computing; electrodynamics; neural nets; nonlinear equations; radial basis function networks; R-functions method; RBF neural networks approach; boundary value problems; eigenvalue problems; electrodynamics problems; nonlinear problems; Atomic measurements; Boundary conditions; Boundary value problems; Electrodynamics; Electronic mail; Finite difference methods; Finite volume methods; Neural networks; Partial differential equations; Physics;
         
        
        
        
            Conference_Titel : 
Physics and Engineering of Microwaves, Millimeter and Submillimeter Waves and Workshop on Terahertz Technologies, 2007. MSMW '07. The Sixth International Kharkov Symposium on
         
        
            Conference_Location : 
Kharkov
         
        
            Print_ISBN : 
1-4244-1237-4
         
        
        
            DOI : 
10.1109/MSMW.2007.4294865