pt.tumba.ngram.svm
Class SVMModel

java.lang.Object
  extended by pt.tumba.ngram.svm.SVMModel

public class SVMModel
extends java.lang.Object

SVMModel encondes a classification model, describing both the model parameters and the Support Vectors.

Author:
Bruno Martins

Field Summary
(package private)  int[] label
          Label of each class (for classification only).
(package private)  int nrClasses
          The number of classes.
(package private)  int[] numSupportVectors
          Number of Support Vectors for each class (for classification only).
(package private)  SVMParameter params
          The parameters of this model.
(package private)  double[] probA
          Pariwise probability information.
(package private)  double[] probB
          Pariwise probability information.
(package private)  double[] rho
          Constants (the bias terms) in decision functions.
(package private)  SVMNode[][] supportVectors
          The Support Vectors.
(package private)  double[][] supportVectorsCoef
          The coefficients for Support Vectors in decision functions.
(package private)  int totalNrSupportVectors
          The total number of Support Vectors.
 
Constructor Summary
SVMModel()
           
 
Method Summary
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

params

SVMParameter params
The parameters of this model.


nrClasses

int nrClasses
The number of classes. (2 in regression/one class svm).


totalNrSupportVectors

int totalNrSupportVectors
The total number of Support Vectors.


supportVectors

SVMNode[][] supportVectors
The Support Vectors.


supportVectorsCoef

double[][] supportVectorsCoef
The coefficients for Support Vectors in decision functions.


rho

double[] rho
Constants (the bias terms) in decision functions.


probA

double[] probA
Pariwise probability information.


probB

double[] probB
Pariwise probability information.


label

int[] label
Label of each class (for classification only).


numSupportVectors

int[] numSupportVectors
Number of Support Vectors for each class (for classification only).

Constructor Detail

SVMModel

public SVMModel()