The Complete SPRLIB & ANNLIB

parzen_mclass

- pattern classification using the Parzen classifier

SYNOPSIS

int parzen_mclass (pattern, S, P)

ARGUMENTS

double *pattern The input pattern which has to be classified, a vector of size [1..n].
double *S A vector, of size [1..m], of smoothing paramaters, one for each class.
double *P A vector, of size [1..m], of apriori class probabilities.

RETURNS

If no error occurs, the function returns the class in which the pattern is classified. In case of an error a negative value is returned.

FUNCTION

This function classifies a pattern pattern using the Parzen classifier. A pattern is classified into the class that has the highest memberships value for the particular pattern. Since this is the multiclass case, S contains the smoothing values for each of the kernels for each class, and P contains the apriori class probabilities.

NOTE

For this function to work the Parzen estimator has to be initialized. This can be done using parzen_init. The variable pattern is in vector format.

SEE ALSO

parzen_classify

This document was generated using api2html on Thu Mar 5 09:00:00 MET DST 1998