moa.evaluation

Interface ClassificationPerformanceEvaluator

All Superinterfaces:

MOAObject, Serializable

All Known Subinterfaces:

RegressionPerformanceEvaluator

All Known Implementing Classes:

BasicClassificationPerformanceEvaluator, BasicConceptDriftPerformanceEvaluator, BasicRegressionPerformanceEvaluator, EWMAClassificationPerformanceEvaluator, FadingFactorClassificationPerformanceEvaluator, MultilabelWindowClassificationPerformanceEvaluator, WindowClassificationPerformanceEvaluator, WindowRegressionPerformanceEvaluator

public interface ClassificationPerformanceEvaluator
extends MOAObject

Interface implemented by learner evaluators to monitor the results of the learning process.

Version:

$Revision: 7 $

Author:

Richard Kirkby (rkirkby@cs.waikato.ac.nz)

Method Summary

Methods

Modifier and Type	Method and Description
void	addResult(weka.core.Instance inst, double[] classVotes) Adds a learning result to this evaluator.
Measurement[]	getPerformanceMeasurements() Gets the current measurements monitored by this evaluator.
void	reset() Resets this evaluator.

Methods inherited from interface moa.MOAObject

copy, getDescription, measureByteSize

Method Detail

reset

void reset()

Resets this evaluator. It must be similar to starting a new evaluator from scratch.

addResult

void addResult(weka.core.Instance inst,
             double[] classVotes)

Adds a learning result to this evaluator.

Parameters:

inst - the instance to be classified

classVotes - an array containing the estimated membership probabilities of the test instance in each class

getPerformanceMeasurements

Measurement[] getPerformanceMeasurements()

Gets the current measurements monitored by this evaluator.

Returns:

an array of measurements monitored by this evaluator