Radial Basis Function Networks with Optimal Kernels

被引：0

作者：

Krzyzak, Adam ^{[1
]}

机构：

[1] Concordia Univ, Dept Comp Sci & Software Engn, Montreal, PQ H3G 1M8, Canada

来源：

2011 IEEE INTERNATIONAL SYMPOSIUM ON INFORMATION THEORY PROCEEDINGS (ISIT) | 2011年

关键词：

Radial basis function networks; optimal kernel; optimal rate of convergence; regression estimation and classification; BASIS FUNCTION NETS; CONVERGENCE-RATES; REGRESSION; DENSITY; RISK;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

We consider nonlinear function estimation using Radial Basis Function Networks. We analytically determine the optimal radial kernel minimizing the Mean Integrated Square Error (MISE) and the optimal MISE rate of convergence. The rates of convergence for various classes of nonlinear functions and input densities are also considered.

引用

页码：860 / 863

页数：4

共 24 条

[1]

[Anonymous], 2004, KERNEL METHODS PATTE

[2]

[Anonymous], 1987, ALGORITHMS APPROXIMA

[3]

Broomhead D. S., 1988, Complex Systems, V2, P321

[4] MEAN INTEGRATED SQUARE ERROR PROPERTIES OF DENSITY ESTIMATES [J].