Title :
Comparison of rates of linear and neural network approximation
Author :
Kurkova, Vera ; Sanguineti, Marcello
Author_Institution :
Inst. of Comput. Sci., Czechoslovak Acad. of Sci., Prague, Czech Republic
Abstract :
We develop some mathematical tools for comparison of rates of fixed versus variable basis function approximation. Using these tools, we describe sets of multivariable functions, for which lower bounds on worst-case errors in approximation by n-dimensional linear subspaces are larger than upper bounds on such errors in approximation by perceptron networks with n hidden units
Keywords :
Hilbert spaces; feedforward neural nets; function approximation; perceptrons; linear approximation; lower bounds; multivariable functions; n-dimensional linear subspaces; neural network approximation; perceptron networks; variable basis function approximation; worst-case errors; Computer errors; Computer networks; Computer science; Electronic mail; Feedforward neural networks; Fourier transforms; Function approximation; Linear approximation; Neural networks; Upper bound;
Conference_Titel :
Neural Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on
Conference_Location :
Como
Print_ISBN :
0-7695-0619-4
DOI :
10.1109/IJCNN.2000.857848
