- Type Parameters:
I -
- All Implemented Interfaces:
- ai.libs.jaicore.basic.algorithm.IAlgorithm<IDataset<I>,IDataset<I>>, ai.libs.jaicore.basic.Cancelable, ai.libs.jaicore.basic.ILoggingCustomizable, java.lang.Iterable<ai.libs.jaicore.basic.algorithm.events.AlgorithmEvent>, java.util.concurrent.Callable<IDataset<I>>, java.util.Iterator<ai.libs.jaicore.basic.algorithm.events.AlgorithmEvent>
public class ClassifierWeightedSampling<I extends IInstance>
extends CaseControlLikeSampling<I>
The idea behind this Sampling method is to weight instances depended on the
way a pilot estimator p classified them. Instances that p classified right
but was unsure contain the most information and are most likely to be chosen.
Instances that p is very sure about and Instances that p is quite sure about
their actual class and classified them falsely, are medium likely to be
chosen. Instances that p is very unsure about their actual class and
classified them falsely are not likely to be chosen. Note that any instance
still has a base probability to be chosen.