gtm_gbf {gtm}R Documentation

Calculates the output of Gaussian basis functions for a given set of input

Description

First the function computes matrix (d_{ij}) of distances between centers of Gaussians given by MU and latent points X. The output values are given by

p(x) = exp(-1/(2* sigma^2)*d_{ij})

Usage

FI = gtm_gbf(MU, sigma, X)

Arguments

MU a M-by-L matrix containing the centers of the basis functions
sigma a scalar giving the standard deviation of the radii-symmetric Gaussian basis functions
X the latent variable sample forming the set of inputs; K-by-L

Value

FI - the matrix of basis functions output values; K-by-(M+1), "+1" for a bias basis function with a fixed value of 1.0

See Also

gtm_lbf

Examples

  
  MU = matrix(c(0,1), ncol = 1)
  sigma = 1
  X = gtm_pts(8)
  FI = gtm_gbf(MU, sigma, X)
  matplot(X, FI, type="l")

[Package gtm version 1.0 Index]