SDP_NORMAL (Low-level) Gaussian model

   PP=SDP_NORMAL(MEAN,COV)
   PP=SDP_NORMAL(MEAN,COV,PRIORS)

 INPUT
  MEAN     SDDATA set with with component means
  COV      Covariance provided as:
             - scalar value (is filled on the covariance diagonal)
             - single covariance matrix (is used for each component)
             - cell array with covariance matrices
             - 3D matrix with individual covariance matrices
               (feat x feat x components)
  PRIORS   (optional) vector with component priors (1 x components)

 OUTPUT
  PP       Pipeline object

 DESCRIPTION
 SDP_NORMAL implements a low-level pipeline constructor for general
 Gaussian model with full covariance matrices for one- or multi-class
 classification.

 SEE ALSO
 SDGAUSS, SDNMEAN, SDLINEAR, SDQUADRATIC, SDMIXTURE