The Complete SPRLIB & ANNLIB

khighley_dataset

- generate samples according to the generalized Highleyman dataset

SYNOPSIS

DATASET *khighley_dataset (NClassA, NClassB, k)

ARGUMENTS

int NClassA The number of samples for class A.
int NClassB The number of samples for class B.
int k The data dimensionality.

RETURNS

The function returns NULL if an error occured, otherwise it returns a valid pointer to a DATASET structure.

FUNCTION

This function generates a k-dimensional dataset with samples distributed according to a generalized Highleyman dataset. Samples of the khighley dataset are drawn as follows: 
Class A :    1    <= k1 < sqrt(k) Mean = 1 Sigma = 0.25
          sqrt(k) <= k  <=   k    Mean = 1 Sigma = 16.0
Class B :    1    <= k1 < sqrt(k) Mean = 1 Sigma = 0.04
          sqrt(k) <= k1 <    k    Mean = 0 Sigma = 16.0

NOTE

The returned pointer points to a dataset which is a LABELSET and LEARNSET (see DATASET-flags).

SEE ALSO

random_khighley_a, random_khighley_b

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