The Complete SPRLIB & ANNLIB

parzen_classify

- two class Parzen classifier

SYNOPSIS

int parzen_classify (pattern, S0, S1, C)

ARGUMENTS

double *pattern The input pattern which has to be classified, a vector of size [1..n].
double S0 The smoothing parameter for class 0.
double S1 The smoothing parameter for class 1.
double C The quotient of the apriori probabilities.

RETURNS

The function returns -1 if an error was detected, if no error was detected it returns the most probable class the pattern pattern belongs to, either 0 or 1.

FUNCTION

This function computes the most problable class a pattern pattern belongs to using a Parzen classifier. The function also needs the apriori probabilities of both classes, C = P(Class1)/P(Class0).

NOTE

This function only works properly if the function parzen_init has been called in advance to initialize the proper variables. The pattern pattern is in vector format.

SEE ALSO

parzen_init, parzen_mclass

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