weka.classifiers
Interface ActiveLearner
- All Known Implementing Classes:
- ActiveDecorate, Fable, QBag, QBoost
- public interface ActiveLearner
Interface to permit a classifier to perform selective sampling. A
classifier that implements this interface can actively select
training examples from a given set of unlabeled examples.
Method Summary |
int[] |
selectInstances(Instances unlabeledActivePool,
int num)
Given a set of unlabeled examples, select a specified number of examples to be labeled. |
selectInstances
public int[] selectInstances(Instances unlabeledActivePool,
int num)
throws java.lang.Exception
- Given a set of unlabeled examples, select a specified number of examples to be labeled.
- Parameters:
unlabeledActivePool
- pool of unlabeled examplesnum
- number of examples to selcted for labeling
- Throws:
java.lang.Exception
- if selective sampling fails