com.structed.models

Class Classifier

java.lang.Object

com.structed.models.Classifier

public class Classifier
extends java.lang.Object

Classifier class is responsible for the training the prediction

Field Summary

Fields

Modifier and Type	Field and Description
ClassifierData	classifierData
java.util.ArrayList<java.lang.Double>	validationCumulativeLoss

Constructor Summary

Constructors

Constructor and Description
Classifier()

Method Summary

Methods

Modifier and Type	Method and Description
Vector	getAvgWeights() a getter for the averaged weights
PredictedLabels	test(Vector W, Example example, int returnAll) predict
Vector	train(Vector W, InstancesContainer trainInstances, java.util.List<java.lang.Double> params, InstancesContainer developInstances, int isAvg, boolean verbose) train the algorithm and return the final weights

Methods inherited from class java.lang.Object

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

Field Detail

classifierData

public ClassifierData classifierData

validationCumulativeLoss

public java.util.ArrayList<java.lang.Double> validationCumulativeLoss

Constructor Detail

Classifier

public Classifier()

Method Detail

train

public Vector train(Vector W,
           InstancesContainer trainInstances,
           java.util.List<java.lang.Double> params,
           InstancesContainer developInstances,
           int isAvg,
           boolean verbose)
             throws java.lang.Exception

train the algorithm and return the final weights

Parameters:

W - model weights

trainInstances - training examples

params - a given parameters for the model

developInstances - development set, if set to null the model skip the validation phase

isAvg - if this parameter set to 1 the function will average the model weights, otherwise it will return only the last one

Returns:

the final model's weights

Throws:

java.lang.Exception

test

public PredictedLabels test(Vector W,
                   Example example,
                   int returnAll)

predict

Parameters:

W - the model weights

example - an example to predict

returnAll - can be used by the inference to return the desired number of examples

Returns:

PredictedLabels object containing the sorted predictions

getAvgWeights

public Vector getAvgWeights()

a getter for the averaged weights

Returns: