magpie.models.classification

Class BaseClassifier

java.lang.Object

magpie.models.BaseModel

magpie.models.classification.BaseClassifier

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, AbstractClassifier, Citable, Commandable, Options, Printable, Savable

Direct Known Subclasses:

CumulantExpansionClassifier, ScikitLearnClassifier, WekaClassifier

public abstract class BaseClassifier
extends BaseModel
implements AbstractClassifier

Base class for all classification models. As such, it provides a constructor and functions to control whether the class variable is treated as discrete or not.

This class also includes a "class cutoff" used with continuous class variables, which is used to convert it to a discrete value. For a predicted class, x, the discrete is floor(x) if rem(x,1) is less than the cutoff and ceil(x) otherwise.

Any models that implements this interface must determine the probability of each class occurring if the class variable is discrete

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected java.lang.String[]	ClassNames Names of classes this model can distinguish between
protected int	NClasses Number of classes to distinguish between

Fields inherited from class magpie.models.BaseModel

AttributeSelector, trained, TrainingStats, validated, ValidationStats

Constructor Summary

Constructors

Constructor and Description
BaseClassifier()

Method Summary

Modifier and Type	Method and Description
BaseClassifier	clone()
java.lang.String[]	getClassNames() Get names of classes this model distinguishes between.
int	getNClasses()
void	run(Dataset TestData) Run a model on provided data.
void	train(Dataset D, boolean b) Train a model on a specified training set and then evaluate performance on the training set, if desired

Methods inherited from class magpie.models.BaseModel

about, crossValidate, crossValidate, crossValidate, done, externallyValidate, getAttributeSelector, getCitations, getFilter, getTrainTime, getValidationMethod, handleSetCommand, isTrained, isValidated, loadState, printCommand, printDescription, printModel_protected, printModel, printModelDescriptionDetails, resetModel, run_protected, runCommand, saveCommand, saveState, setAttributeSelector, setComponent, setFilter, train_protected, train

Methods inherited from class java.lang.Object

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

Methods inherited from interface magpie.utility.interfaces.Options

printUsage, setOptions

Field Detail

NClasses

protected int NClasses

Number of classes to distinguish between

ClassNames

protected java.lang.String[] ClassNames

Names of classes this model can distinguish between

Constructor Detail

BaseClassifier

public BaseClassifier()

Method Detail

clone

public BaseClassifier clone()

Overrides:

clone in class BaseModel

getNClasses

public int getNClasses()

Specified by:

getNClasses in interface AbstractClassifier

Returns:

Number of classes model will distinguish between

getClassNames

public java.lang.String[] getClassNames()

Description copied from interface: AbstractClassifier

Get names of classes this model distinguishes between.

Specified by:

getClassNames in interface AbstractClassifier

Returns:

Names of classes

train

public void train(Dataset D,
                  boolean b)

Description copied from class: BaseModel

Train a model on a specified training set and then evaluate performance on the training set, if desired

Overrides:

train in class BaseModel

Parameters:

D - Dataset to use for training

b - Whether to record training statistics

run

public void run(Dataset TestData)

Description copied from class: BaseModel

Run a model on provided data. Results will be stored as the predicted class variable.

Overrides:

run in class BaseModel

Parameters:

TestData - Dataset to evaluate.