NMC

Nearest Mean Classifier

    W = NMC(A)
    W = A*NMC

Input
A	Dataset

Output
W	Nearest Mean Classifier

Description

Computation of the nearest mean classifier between the classes in the  dataset A. The use of soft labels is supported. Prior probabilities are  not used.

The difference with NMSC is that NMSC is based on an assumption of normal  distributions and thereby automatically scales the features and is  sensitive to class priors. NMC is a plain nearest mean classifier for  which the assigned classes are are sensitive to feature scaling and  unsensitive to class priors.

The estimated class confidences by B*NMC(A), however, are based on  assumed spherical gaussian distributions of the same size.

As NMC is a simple linear classifier, a non-linear combiner might give a  significant performance improvement in multi-dimensional problems, e.g.  by W = A*(NMC*QDC([],[],1e-6)).

See also

datasets, mappings, nmsc, ldc, fisherc, qdc, udc,