Title :
On MISE convergence rates of radial basis functions networks
Author :
Krzyiak, A. ; Niemann, H.
Author_Institution :
Dept. of Comput. Sci., Concordia Univ., Montreal, Que., Canada
Abstract :
We investigate mean integrated squared error (MISE) convergence and the rate of convergence of a class of RBF networks. This paper extends L2 convergence results for radial basis nets given in Xu, Krzyzak and Yuille (1994). We obtain tighter bounds and under milder conditions for MISE than those obtained previously for L2 error
Keywords :
approximation theory; convergence; feedforward neural nets; L2 convergence results; MISE convergence rates; mean integrated squared error; radial basis functions networks; Approximation error; Collaboration; Computer errors; Computer science; Convergence; H infinity control; Kernel; Neural networks; Radial basis function networks; Upper bound;
Conference_Titel :
Neural Networks, 1996., IEEE International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-3210-5
DOI :
10.1109/ICNN.1996.548897
