Serialized Form


Package jml.classification

Class jml.classification.AdaBoost extends Classifier implements Serializable

serialVersionUID: 1100546985050582205L

Serialized Fields

T

int T
Number of iterations, or the number of weak classifiers.


weakClassifiers

Classifier[] weakClassifiers
The sequence of weak classifiers during training.


alphas

double[] alphas
Weights on the outputs of the trained weak classifiers.

Class jml.classification.Classifier extends java.lang.Object implements Serializable

serialVersionUID: 7315629872244004590L

Serialized Fields

nClass

int nClass
Number of classes.


nFeature

int nFeature
Number of features, without bias dummy features, i.e., for SVM.


nExample

int nExample
Number of samples.


X

org.apache.commons.math.linear.RealMatrix X
Training data matrix (nFeature x nExample), each column is a feature vector. The data matrix should not include bias dummy features.


Y

org.apache.commons.math.linear.RealMatrix Y
Label matrix for training (nExample x nClass). Y_{i,k} = 1 if x_i belongs to class k, and 0 otherwise.


labelIDs

int[] labelIDs
LabelID array for training data, starting from 0. The label ID array for the training data is latent, and we don't need to know them. They are only meaningful for reconstructing the integer labels by using IDLabelMap structure.


labels

int[] labels
Label array for training data with original integer code.


W

org.apache.commons.math.linear.RealMatrix W
Projection matrix (nFeature x nClass), column i is the projector for class i.


epsilon

double epsilon
Convergence tolerance.


IDLabelMap

int[] IDLabelMap
An ID to integer label mapping array. IDs start from 0.

Class jml.classification.LogisticRegressionMCBoundConstrainedPLBFGS extends Classifier implements Serializable

serialVersionUID: 8822935620181763793L

Class jml.classification.LogisticRegressionMCGradientDescent extends Classifier implements Serializable

serialVersionUID: 795431672777823665L

Class jml.classification.LogisticRegressionMCLBFGS extends Classifier implements Serializable

serialVersionUID: 337850590370432119L

Serialized Fields

model

Model model

Class jml.classification.LogisticRegressionMCLBFGS_Ori extends Classifier implements Serializable

serialVersionUID: -2027322010064300037L

Class jml.classification.LogisticRegressionMCNonlinearConjugateGradient extends Classifier implements Serializable

serialVersionUID: 3852966108762824837L

Class jml.classification.LogisticRegressionMCNonnegativePLBFGS extends Classifier implements Serializable

serialVersionUID: -2215716367007018997L

Class jml.classification.MaxEnt extends Classifier implements Serializable

serialVersionUID: -316339495680314422L

Serialized Fields

F

org.apache.commons.math.linear.RealMatrix[] F
Feature matrix array. F[n] is the feature matrix for the n-th sample, where F[n][i][k] is the i-th feature value on the k-th class for the n-th sample.

Class jml.classification.MultiClassSVM extends Classifier implements Serializable

serialVersionUID: 6720660840680032348L

Serialized Fields

problem

de.bwaldvogel.liblinear.Problem problem
A Problem object for this MCSVM classifier.


features

de.bwaldvogel.liblinear.Feature[][] features
Feature 2D array, indices start from 1.


C

double C
Parameter for loss term of linear multi-class SVM.


eps

double eps
Convergence tolerance.


parameter

de.bwaldvogel.liblinear.Parameter parameter
A Parameter instance for linear multi-class SVM.


bias

double bias
Dummy feature, aiming to remove the equality constraint for the dual problem of SVM.


model

de.bwaldvogel.liblinear.Model model
SVM model.