wolffd@0: wolffd@0:
wolffd@0:wolffd@0: cov = gpcovar(net, x) wolffd@0: [cov, covf] = gpcovar(net, x) wolffd@0:wolffd@0: wolffd@0: wolffd@0:
cov = gpcovar(net, x)
takes
wolffd@0: a Gaussian Process data structure net
together with
wolffd@0: a matrix x
of input vectors, and computes the covariance
wolffd@0: matrix cov
. The inverse of this matrix is used when calculating
wolffd@0: the mean and variance of the predictions made by net
.
wolffd@0:
wolffd@0:
[cov, covf] = gpcovar(net, x)
also generates the covariance
wolffd@0: matrix due to the covariance function specified by net.covarfn
wolffd@0: as calculated by gpcovarf
.
wolffd@0:
wolffd@0:
x
and is then
wolffd@0: passed to gpfwd
so that predictions (with mean ytest
and
wolffd@0: variance sigsq
) can be made for the test inputs
wolffd@0: xtest
.
wolffd@0: wolffd@0: wolffd@0: cninv = inv(gpcovar(net, x)); wolffd@0: [ytest, sigsq] = gpfwd(net, xtest, cninv); wolffd@0:wolffd@0: wolffd@0: wolffd@0:
gp
, gppak
, gpunpak
, gpcovarp
, gpcovarf
, gpfwd
, gperr
, gpgrad
Copyright (c) Ian T Nabney (1996-9) wolffd@0: wolffd@0: wolffd@0: wolffd@0: