SciLib.Neuro
Class Classify

java.lang.Object
  SciLib.Neuro.Classify

public final class Classify
extends java.lang.Object

This class defines methods used to solve a classification problem.

Constructor Summary

Classify()

Method Summary

static int	checkValidClassify(vector y) Classify a pattern y by the check valid method
static void	chooseKRowsRandomly(matrix m, integerVector rowIndexes, int k) Choose randomly k rows of the matrix m.
static int	classify(vector y, int type) Classify a pattern y
static integerVector[]	kMeanClustering(int k, matrix m, matrix cs) Compute a matrix that contains k clustering patterns
static int	nearest(vector x, matrix m) Find the row index of the matrix m that has minimum distance to x.
static int	nearestClassify(vector y) Classify a pattern y by the nearest method

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Classify

public Classify()

Method Detail

classify

public static int classify(vector y,
                           int type)

Classify a pattern y

Parameters:

y - A vector, the pattern

type - An integer value, 0 for the nearest neighbor method 1 for the CheckValid method

Returns:

an integer

nearestClassify

public static int nearestClassify(vector y)

Classify a pattern y by the nearest method

Parameters:

y - A vector, the pattern

Returns:

an integer

checkValidClassify

public static int checkValidClassify(vector y)

Classify a pattern y by the check valid method

Parameters:

y - A vector, the pattern

Returns:

an integer

nearest

public static int nearest(vector x,
                          matrix m)

Find the row index of the matrix m that has minimum distance to x.

Parameters:

x - A vector

m - A matrix

Returns:

an integer

chooseKRowsRandomly

public static void chooseKRowsRandomly(matrix m,
                                       integerVector rowIndexes,
                                       int k)

Choose randomly k rows of the matrix m. The result saved in rowIndexes vector

Parameters:

m - A matrix

rowIndexes - An integerVector

k - An integer

kMeanClustering

public static integerVector[] kMeanClustering(int k,
                                              matrix m,
                                              matrix cs)

Compute a matrix that contains k clustering patterns

Parameters:

k - An integer

m - A matrix

cs - A matrix