The Complete SPRLIB & ANNLIB

parzen_dataset_class

- classify a sample, given a dataset, using the parzen estimator

SYNOPSIS

int parzen_dataset_class (dset, samp, S, P, NCls)

ARGUMENTS

DATASET *dset The dataset used to train the parzen classifier.
SAMPLE *samp The sample to be classified.
double *S The smoothing paramaters per class, a vector of size [1..NCls].
double *P The apriori probabilites per class, a vector of size [1..NCls].
int NCls The number of classes.

RETURNS

The function returns the class label of the most probable class the sample samp belongs to.

FUNCTION

This function calculates the class label of the class to which the sample samp belongs, i.e. the class that results in the highest aposteriori probability. The dataset dset is used to train the parzen classifier, i.e. to compute a kernel for each class. In order to do this, it needs the smoothing paramaters, S, and the apriori probabilities, P, per class.

NOTE

The variables S and P are NOT in vector format but are indexed from 0. The dataset dest must be a LABELSET, see DATASET-flags.

SEE ALSO

parzen_probability, parzen_mclass, parzen_best_s, parzen_dset_best_s, find_points(1)

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