ai.libs.jaicore.ml.activelearning

Class PoolBasedUncertaintySamplingStrategy<T,I extends ILabeledInstance,D extends IDataset<I>>

java.lang.Object

ai.libs.jaicore.ml.activelearning.PoolBasedUncertaintySamplingStrategy<T,I,D>

Type Parameters:

T - TARGET

I - The class of instances

D - The class of the dataset

All Implemented Interfaces:

ISelectiveSamplingStrategy<I>

public class PoolBasedUncertaintySamplingStrategy<T,I extends ILabeledInstance,D extends IDataset<I>>
extends java.lang.Object
implements ISelectiveSamplingStrategy<I>

A simple pool-based uncertainty sampling strategy, which assesses certainty for all instances in the pool and picks the instance with least certainty for the next query.

Constructor Summary

Constructors

Constructor and Description
PoolBasedUncertaintySamplingStrategy(ICertaintyProvider<T,I,D> certaintyProivder, IActiveLearningPoolProvider<I> poolProvider)

Method Summary

Modifier and Type	Method and Description
I	nextQueryInstance() Chooses the IInstance to query next.

Methods inherited from class java.lang.Object

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

Constructor Detail

PoolBasedUncertaintySamplingStrategy

public PoolBasedUncertaintySamplingStrategy(ICertaintyProvider<T,I,D> certaintyProivder,
                                            IActiveLearningPoolProvider<I> poolProvider)

Method Detail

nextQueryInstance

public I nextQueryInstance()

Description copied from interface: ISelectiveSamplingStrategy

Chooses the IInstance to query next.

Specified by:

nextQueryInstance in interface ISelectiveSamplingStrategy<I extends ILabeledInstance>

Returns:

IInstance to query next.