PARZENDC

Parzen density based classifier

   [W,H] = PARZENDC(A)
   W = PARZENDC(A,H)

Input
A	Dataset
H	Smoothing parameters (optional; default: estimated from A for each class)

Output
W	Trained Parzen classifier
H	Smoothing parameters, estimated from the data

Description

For each of the classes in the dataset A, a Parzen density is estimated  using PARZENML. For each class, a feature normalisation on variance is  included in the procedure. As a result, the Parzen density estimate uses  different smoothing parameters for each class and each feature.

If a set of smoothing parameters H is specified, no learning is performed,  only the classifier W is produced. H should have the size of [C x K] if  A has C classes and K features. If the size of H is [1 x K] or [C x 1],  or [1 x 1], then identical values are assumed for all the classes and/or  features.

The densities for the points of a dataset B can be found by D = B*W.  D is an [M x C] dataset, if B has M objects.

Example(s)

See also

datasets, mappings, parzenc, parzen_map, parzenml,