PRTools contents

PRTools manual

normal_map

NORMAL_MAP

Map a dataset on normal-density classifiers or mappings

    F = NORMAL_MAP(A,W)

Input
 A Dataset
 W Mapping

Output
 F Density estimation for classes in A

Description

Maps the dataset A by the normal density based classifier or mapping W.  For each object in A, F returns the densities for each of the classes or  distributions stored in W. For classifiers, the densities are weighted  by the class prior probabilities. This routine is automatically called for  computing A*W if W is a normal density based classifier or a mapping.

Use W = LOGDENS(W) (or W = W*LOGDENS) if absolute densities are not  needed and a more accurate posterior probability is desired.

See also

mappings, datasets, qdc, udc, ldc, gaussm, logdens,

PRTools contents

PRTools manual